1 | Session Title | Session Description | Organizer(s) Name | Organizer(s) Affiliation | Chair Name | Chair Affiliation | Speakers |
---|---|---|---|---|---|---|---|
2 | Advanced methods for data with heterogeneity | This section mainly focus on the topic of advanced methods for data with heterogeneity. Topics will cover offline and online change point detection and high dimensional quantile regressions. | Bin; Yufeng Liu; Liu | Fudan University; UNC-Chapel Hill | Yufeng Liu | UNC-Chapel Hill | Haojie Ren, Xianru Wang, Bin Liu, Wenqi Lu |
3 | Advanced Methods in Large-scale Statistical Analysis | This session will focus on advanced methods in the field of large-scale statistical analysis. We will delve into new techniques and innovative approaches required for handling massive datasets and complex models. Topics for discussion include the application of machine learning in large-scale datasets, distributed computing, and optimization algorithms. We invite professionals with extensive experience in statistics and related fields to share their research findings and insights. The session aims to provide attendees with an in-depth understanding of the latest developments in the field of large-scale statistical analysis, fostering knowledge exchange and potential collaborations for the future. | Shuyuan Wu | Shanghai University of Finance and Economics | Shuyuan Wu | Shanghai University of Finance and Economics | Haobo Qi, Ziqian Lin, Jiyuan Tu, Shuyuan Wu |
4 | Advanced Statistical Approaches for Complex Data | In this session we will discuss several advanced statistical approaches for complex data such as rerandomization and covariate adjustment in split-plot designs, prior information-assisted integrative analysis of multiple datasets, model-free joint variable selection via importance weighting, and estimating large-scale spatial autoregression model by randomly approximated gradient descent. | Yufeng Liu | University of North Carolina at Chapel Hill | Xiaoling Lu | Renmin University of China | Hanzhong Liu, Feifei Wang, Yiwei Fan, Yuan Gao |
5 | Advanced statistical methods for bulk and single-cell genomics | In the rapidly evolving landscape of genomics, the integration of advanced statistical methods has become imperative for extracting meaningful insights from high-dimensional data. This session on "Advanced Statistical Methods for Bulk and Single-Cell Genomics" aims to bring together statisticians and biostatisticians to explore cutting-edge methodologies and applications in the field. Advancements in technology have enabled the acquisition of vast amounts of genomic data, both at the bulk and single-cell levels. This session will delve into the challenges posed by such data richness and showcase innovative statistical approaches designed to unravel the complexities inherent in genomics studies. From understanding cellular heterogeneity to identifying key molecular signatures, the session will span a spectrum of topics crucial for the analysis of RNA sequencing, spatial transcriptomics, and TCR sequencing data. | Wei Vivian Li | University of California, Riverside | Wei Vivian Li | University of California, Riverside | Ying Ma, Hongkai Ji, Jian Yang, Jessica Li |
6 | Advanced Statistical Methods for Medical Data Analysis | This session focuses on the application of state-of-the-art statistical and machine learning methods for the analysis of various medical data. The content encompasses medical imaging data, clinical trials, and diagnostic data. The audiences cab gain insights into the latest methodologies driving advancements in healthcare analytics. | Jing Zhou | Renmin University of China | Jing Zhou | Renmin University of China | Ben Wu, Guangyu Yang, Mengjie Fang, Qianhan Zeng |
7 | Advanced statistical methods in the analysis of high-dimensional and complex data | In our session, we will present advanced statistical methods for high-dimensional and complex data in multiple application fields, such as genomics and insurance. The proposed methods are based on advanced transfer learning, penalization, and Bayesian techniques, which have solid theoretical ground. | Mengyun Wu | Shanghai University of Finance & Economics | Mengyun Wu | Shanghai University of Finance & Economics | Jingsi Min, Xiaonan Hu, Yaqing Xu, Hao Mei |
8 | Advancements in Statistical Methods for Diverse Applications | This session presents some new studies on statistical learning for high dimensional and large scale data. | yanyan Liu | Wuhan University | Wen Su | Department of Biostatistics, City University of Hong Kong | wen Su, Jing Zhang, Xiaochao Xia, Lican Kang |
9 | Advancements in Statistical Testing, Estimation, and Modeling across Diverse Data Structures | This session showcases cutting-edge methodologies and applications designed to navigate the complexities inherent in diverse data structures. From exploring novel testing techniques to inference methodologies and modeling approaches, this session offers some of the latest advancements in statistical analysis. The talks featured in this session cover a broad spectrum of statistical domains: * "Two-Sample Tests Based on Data Depth" introduces innovative approaches leveraging data depth for robust and nuanced two-sample testing. Discover how these methods handle complex data structures, ensuring reliability and accuracy in statistical inference. * "Estimation and Inference for Fixed Center Effects on Panel Count Data" delves into the intricacies of panel count data, offering insights into estimation techniques and inference methods tailored to uncover hidden patterns and effects within such structured data. * "Fitting Survival Models Using Auxiliary Summary Information" explores the integration of auxiliary summary information into survival models, illuminating how these models can be enhanced and refined for more accurate predictions and insights. * "Consistent Semiparametric Tests for Lorenz Dominance Based on the Density Ratio Model" introduces innovative approaches to assess income distribution inequalities. Understand how consistent semiparametric tests based on the density ratio model can unveil significant differences in income distributions. | Weixin Yao | University of California, Riverside | Weiwei Zhuang | Zhejiang Gongshang University | Chifeng Shen, Weiwei Zhuang, Xiaoguang Wang, Weiwei Wang |
10 | Advances in Causal Inference: Methods and Applications | This session highlights cutting-edge research in causal inference, focusing on methods applicable to critical and complex real-world scenarios. It aims to introduce reliable and effective methodologies, emphasizing their pivotal role in practical applications. Our speakers will delve into novel approaches to data integration across diverse causal inference scenarios, establish robust and efficient statistical methodologies, and apply advanced statistical tools to address significant real-world challenges. | Yuqian Zhang | Institute of Statistics and Big Data, Renmin University of China | Yuqian Zhang | Institute of Statistics and Big Data, Renmin University of China | Xiaojie Mao, Wei Li, Tuo Lin, Xiao Wu |
11 | Advances in Kernel Methods and Nonparametric Inference for Complex Data | This session presents a series of presentations centered around the application of kernel methods and nonparametric techniques in statistical analysis, particularly for high-dimensional data and streaming data. The session begins with a novel nonparametric multiple sample testing approach, then moves on to scalable non-parametric inference for large-scale streaming data. Subsequent discussions will explore the optimality of kernel regression and kernel classification, with a focus on large-dimensional spaces. This session is designed for audiences interested in recent developments in kernel methods and nonparametric statistical analysis. | Dongming Huang | National University of Singapore | Dongming Huang | National University of Singapore | Jin-Ting Zhang, Meimei Liu, Weihao Lu, Dongming Huang |
12 | Advances in Statistical Learning Inference with High Dimensional Data | Over the past decades, statistical learning and high demensional data have been coevolving and feeding each other. The breadth, complexity, and rapidly expanding size of high dimensional data have stimulated the development of novel statistical learning methods. And application of these methods to high dimensional data have led to many scientific discoveries and practical solutions. Substantial developments have been taking place in statistical learning and algorithms development, driven largely by a wide range of applications in many fields such as genomics, economics, genetics and health record data. In particular, significant advances have been made in the areas of deep learning methods and optimization methods for inference with complex high dimensional data such as genetics and economics data. By computational analyses using statistical learning or nonconvex optimization approaches, one can learn characteristics within samples and identify key features for individual classification and prediction. In this session, we discuss some of the newly arising issues from analyzing high dimensional data to explore key ideas for statistical learning and algorithms with large and complicated modern data generated from various application areas. | Feifei Xiao | University of Florida | Feifei Xiao | University of Florida | Yue Niu, Yeqing Zhou, Feifei Xiao, Qing Cheng |
13 | Advances in statistical methods for -omics data analysis | Recent and rapid advent of high-throughput sequencing technologies has significantly advanced the realm of -omics research. In this proposed invited session, we have invited a group of distinguished statisticians who have impactfully contributed to the development of statistical methods and computational algorithms tailored for -omics data analysis. The lineup includes two speakers based in the United States and two in China, offering a diverse perspective. The proposed topics will span across network analysis, microbiome studies, and spatial transcriptomics data analysis, providing attendees with valuable insights into cutting-edge advancements in these areas. | Hongmei Jiang | Northwestern University | Hongmei Jiang | Northwestern University | Wenxuan Zhong, Yuan Jiang, Tao Wang, Yang Xu |
14 | Advances in theories and methods for dependent data analysis | Dependent data are ubiquitous in such diverse fields as finance, astronomy, seismology, geography, econometrics, medical science, and so on. In this session, some novel theories and methods for analyzing dependent data drawn from time series, stochastic processes and high-dimensional processes will be introduced. | Shibin Zhang | Shanghai Normal University | Shibin Zhang | Shanghai Normal University | Guangying Liu, Yuping Song, Hanchao Wang, Shibin Zhang |
15 | Advancing Interpretability in Statistical and Machine Learning for Transcriptomics and Genomics | This session emphasizes the significance of interpretability in biological research. Our speakers will unveil innovative statistical approaches to develop interpretable models tailored for high-throughput transcriptomic and genomic data. These methods aim to enhance the understanding and applicability of complex biological data. | Jun Li | University of Notre Dame | Jun Li | University of Notre Dame | Sha Cao, Hui Jiang, Jiping Wang, Jun Li |
16 | AI Frontiers in Clinical Data Analysis: Innovations and Applications | Session designed to explore the cutting-edge advancements and applications of artificial intelligence in the field of statistics/clinical data analysis. This session will offer a comprehensive overview of the latest trends, challenges, and innovations in the integration of AI into clinical research. It aims to provide a platform for professionals and researchers to engage in in-depth discussions, share insights, and explore the future potential of AI in transforming clinical data analysis and healthcare outcomes. | Jingwen Jia | Astellas | Jingwen Jia | Astellas | Ming ZHU, Tianlun GU, Bogong ZHU, Haoda FU |
17 | AI in Bioinformatics and Biomedical Applications | Zhaojun Wang | School of Statistics and Data Science, Nankai University | Gang Hu | Nankai University | Zhengling Peng, Meng Wang, Kui Wang, Xiangjie Li | |
18 | Application of causal inference approaches in global drug development | Causal thinking and related inference methods are gaining increasing prominence in global drug development. Causal inference methods for estimating effects of treatments or other interventions on health outcomes have seen rapid and extensive developments in recent years. Tasks for making causal inferences range from estimating average treatment effects, to conditional average treatment effects, to obtaining individualized predictions under different treatment choices. Causal inference generally requires expert knowledge and untestable assumptions about the causal network linking treatment, outcome, and other variables. In this session we build upon these recent developments by examining how causal inference provides a convenient mathematical language and tools to formally establish causal relationships between, e.g., drug and effect. We will show that causal inference methods can be used in many drug development settings, including inference for a principal stratum estimand; leveraging of real-world data and external trial data; and estimation of marginal structural model with unmeasured confounders. | Jiawei Wei | Novartis Institutes for Biomedical Research Co., Shanghai, China | Ziqiang Zhao | Novartis Institutes for Biomedical Research Co., Shanghai, China | Yongming Qu, Ying Wu, Yuan Tian, Yingchun Zhou |
19 | Advanced Statistical Learning For Complicated Gene Data | Since complex disease that strongly connects to genes is one of the most important factors for human health, it is very important to study the relationship between complex diseases and genes. The occurrence and development of complex disease are not only governed by multiple genes and the interaction between genes simultaneously, but also effected by individual trails, environmental factors, and the interaction between genes and environmental factor. The complexity of the network relationship between complex diseases and genes with the huge size of genes brings big challenges for the statistical modelling, estimation and statistical inference. This session will discuss complex networks to establish high-dimensional statistical models. | Xu Liu | Shanghai University of Finance and Economics | Xingjie Shi | East China Normal University | Xingjie Shi, Mengyun Wu, Shucong Zhang, Xiao Zhang |
20 | Big data and nonparametric time series theory and applications | This session is about to report some new developments on big data modeling and nonparametric time series theory and applications. | Chaohua Dong | Zhongnan University of Economics and Law | Feipeng Zhang | Xi'an Jiaotong University | Li Chen, Chen Zhou, Xinqi Wu, Feipeng Zhang |
21 | Biostatistical optimization on multi-omics | The session focus on the biostatistical optimization on multi-omics. We have invited 4 distinguished speakers and they will present the computation methods on multi-omics data and identify the important biological functions. Specifically, the first speaker comes from Beijing Institute of Genomics, CAS and he will give a talk about the role of alternative splicing; The second speaker comes from Shandong University and he will give a talk on identifying the phenotype-associated subpopulations by integrating bulk and single-cell sequencing data; The third speakers comes from Wuhan University and she will give a talk about identifying condition-specific gene expression patterns through aligning multiple spatial transcriptomic datasets. The fourth speakers comes from Xi'an Jiaotong University and he will give a talk about Chinese pan-genome construction and analysis. | Fuke Wu | Huazhong University of Science and Technology | Meng Zou | Huazhong University of Science and Technology | Zhaoqi Liu, Duanchen Sun, Lihua Zhang, Xiaofei Yang |
22 | Causal Inference and Semiparametric Learning | Causal Inference and Semiparametric Learning represent cutting-edge methodologies at the intersection of statistics and machine learning. This session presents a diverse spectrum of research at the statistical frontier. The first report focuses on distribution-free prediction intervals which is crucial for causal inference under covariate shift, while the second contributes by proposing approaches for causal inference with interference in dyadic data. The third report offers a pioneering approach to root-n consistent semiparametric learning in the high-dimensional landscape. Finally, the session concludes with an innovative semiparametric instrumented difference-in-differences approach for policy learning. Together, these reports show the evolving landscape of statistical methodologies, demonstrating the applicability of causal inference and semiparametric learning across various domains. | Wei Li | Renmin University of China | Wei Li | Renmin University of China | Yukun Liu, Wang Miao, Yuhao Wang, Pan Zhao |
23 | Challenges in random field-based modelling and computations in spatial statistics | The goal of the proposed session is to present both theoretical and computational challenges in spatial statistical modeling and inference. Recent developments in theory and applications of random fields are presented to deal with complex dependence structure of multivariate data in space and time. Optimal low-rank approximations are studied to achieve mitigating the computational burden without sacrificing too much information. Some strategical approaches are also proposed to handle non-stationarity in spatial statistical inference. | Juan Du | Kansas State University | Zhengyuan Zhu | Iowa State University | Hao Zhang, Chunsheng Ma, Dan Cheng, Weixing Song |
24 | Complex Data: Methodology and Application | The rapid evolution of technology enables scientists to measure increasingly high-dimensional and complex data, such as large Omics databases, high-frequency time series, 3D images, and among others. This trend poses enormous challenges to data analysis and model inference. Every day researchers push the boundary of what and how we are able to measure information in an unprecedented way in human history. Proper data understanding is key to reaching meaningful conclusions and improving personalized clinical treatment. Accounting for the volume and complexity requires the development of novel statistical models and computational tools, opening up a fascinating and fast-growing area of new frontier in statistics. Bayesian statistics and machine learning methods provide the basic tools to extract knowledge from data, to quantify the associated uncertainty, and to predict or make decisions accounting for such uncertainty. The computing capacity now available has made the use of these methods practicable, and such methods are becoming a vital component in uncovering sparse signals and their interactions. Therefore, I have organized this Invited Session and included Four speakers from the United States and mainland China who respectively work on imaging analysis, diffusion model, Bayesian inference, and graphical model. Speakers will present their theoretical and methodological works on modeling high dimensional data with complex structures to promote informing decisions and driving discovery. This session will foster collaboration opportunities between junior faculty and senior faculty across research institutions. The meetings and discussions will attract a lot of interest and invoke the development of new methodology to solve complex problems in real applications. Significance: In an era of big data, understanding and addressing the challenges of high-dimensional datasets are crucial. This session will offer a platform for researchers to share methodologies, discuss emerging trends, and foster collaboration in advancing statistical techniques in this rapidly evolving field. We believe that this proposed session aligns well with the interests of ICSA attendees, providing valuable insights and encouraging dialogue among statisticians, researchers, and practitioners. | Liangliang Zhang | Case Western Reserve University | Liangliang Zhang | Case Western Reserve University | Weihong Guo, Jiangtao Duan, Chenyu Liu, Ziqi Chen |
25 | Complex Medical Data Analysis | Bring four speakers to present the methods they developed for specific medical datasets recently at the 2024 ICSA China. | Hua Liang | George Washington University | Hua Liang | George Washington University | Hongyuan Cao, Xinmin Li, Peifeng Ruan, Min Tan |
26 | Complex statistical inference for large-scale datasets | Complex statistical inference for large-scale datasets is an emerging issue that aims to tackle the challenges encountered when analyzing massive and complex datasets. In the era of big data, traditional statistical methods face difficulties in terms of computational efficiency and biases. In this regard, subsampling methods have demonstrated significant value. By reducing the dataset size, subsampling overcomes the computational burdens implicated in big data analysis while yielding precise and dependable estimates. This session involves newly proposed methods and theoretical results related to complex statistical inference for large-scale datasets and development of advanced statistical algorithms that can handle the unique features of big data such as high dimensionality, heterogeneity, and sparsity. | Hansheng Wang | Peking University | Yingqiu Zhu | University of International Business and Economics | Wei Hu, Xuetong Li, Yingqiu Zhu, Jun Yu |
27 | Computational and methodological statistics | Computational and methodological statistics constitute a dynamic field at the intersection of statistical theory and modern computing techniques. This discipline focuses on developing and implementing advanced computational methods to address complex statistical problems and extract meaningful insights from large and intricate datasets. By leveraging computational tools, such as algorithms, simulations, and machine learning approaches, statisticians in this field aim to enhance the efficiency, accuracy, and scalability of statistical analyses. Moreover, computational and methodological statistics actively contribute to the evolution of statistical methodologies, exploring innovative ways to handle diverse data types, account for non-standard distributions, and improve the reliability of inference methods. This interdisciplinary approach is particularly crucial in the era of big data, where traditional statistical methods may face challenges in processing and interpreting vast amounts of information. As technology continues to advance, the synergy between computational techniques and statistical methodologies remains instrumental in unlocking new frontiers for data analysis and decision-making across various domains. | Zhigen Zhao | Temple University | Zhigen Zhao | Temple University | Xi Luo, Zhihua Su, Yue Shi, Ying Liao |
28 | Computational Criminal Law | As China continues to make headway with its smart court development project, the smart trial supporting system has been widely used in China’s judicial trial practice. China’s public are paying more attention and developing higher expectations for high-standard and fair judicial practice. This session expects to use statistical methods to provide useful references and a foundation for achieving judicial justice and promoting the development of criminal legislation and theory. | Ke Xu | University of International Business and Economics | Ke Xu | University of International Business and Economics | Xin Liu, Jiaxin Shi, Fang Wang, Ke Xu |
29 | Designs for Computer Experiments | This session includes four speeches on Designs for Computer Experiments. | Min-Qian Liu | Nankai University | Yongdao Zhou | Nankai University | Qian Xiao, Ye Tian, Wenlong Li, Guanzhou Chen |
30 | Early Clinical Trial development and biomarker. | This session is to have an explore on early clinical trials with biomarker studies. Phase I first-in-human (FIH) clinical trials involve initial administration of an investigational product to humans to explore safety and tolerability, pharmacokinetics (PK), pharmacodynamics (PD) as well as early evaluation of drug mechanism (proof of mechanism, POM) before conducting the late-stage trials. The trial scope covers FIH trials, drug-drug interactions, food effects, POM, special populations etc. Trial design is flexible including open-label and baseline-controlled, randomized and double-blind, parallel/crossover design. This session will have four topics including FIH, Drug and drug interaction, Special population study, and biomarker analysis. Each topic will include several real cases with study design, data analysis and impact on drug development. | Na Yang | Sanofi | Amy Zhao | Sanofi | catherine Wang, Yanzhen Wu, Kong Xin, Meiyue Wang |
31 | Efficient analysis in statistics and related fields | The session consists of 4 talks, all related to efficient analysis in some important statistical or related problems. More specifically, the first talk is to develop a sequential algorithm to support efficient quantile regression analysis for massive stream data. The second talk is to propose more efficient estimation for the difference of two quantiles that accounts for the dependence between paired samples. The third talk is to introduce an unsupervised feature selection framework for more efficient analysis of large-scale bioinformatic data. The last talk is to develop an efficient estimator for the parameters in a homogeneous partially linear model without the independence aforementioned. | Tiejun Tong | Hong Kong Baptist University | Zhijian Li | BNU-HKBU United International College | Nan Lin, Yichuan Zhao, Wanjun Ning, Zhijian Li |
32 | Emerging Theory and Methods in Data Science | As technology advances and the volume of available data continues to grow, data science is constantly adapting and incorporating new theories and methodologies. This encompasses a wide range of interdisciplinary approaches, including statistics, computer science, machine learning, and domain-specific knowledge. The emergence of new theories in data science involves the development of conceptual frameworks to understand and model complex phenomena. This may include advancements in statistical modeling, data visualization, and the integration of artificial intelligence and machine learning algorithms. The theories contribute to a deeper understanding of data patterns, relationships, and underlying structures, enabling more accurate predictions and insights. | Ping Ma | UGA | Cheng Meng | RUC | Chuanhai Liu, Ke Deng, Yichao Wu, Changliang Zou |
33 | Empirical Research Methods in Business Applications | The four presentations in this session utilized multivariate statistical analysis and / or structural equation modeling methods to business applications ranging from 1) Studying the impact of digital transformation to agricultural supply chain resilience and responsiveness by Dr. Xiaofei Zhao, 2) Testing the hypotheses about the adoption of blockchain technology for information sharing and humanitarian supply chain operations by Dr. Yinghong Dong, 3) Research about the relationships of uncertainty to humanitarian supply chain resilience and responsiveness by Dr. Yingying Sun and Dr. Ping Wang, and 4) Internet of Things (IoT)-Enabled accountability in source separation of household waste for a circular economy in China by Dr. Bill Wang. 1) Dr. Xiaofei Zhao – Zhongnan University of Nationality The impact of digital transformation to agricultural supply chain resilience and responsiveness 2) Dr. Yinghong Dong – Hainan University A Study of adoption of blockchain technology for information sharing and humanitarian supply chain operations 3) Dr. Yingying Sun, China Renmin University and Dr. Ping Wang, James Madison University "The Impact of Preparation on Humanitarian Response: A Moderated Mediation Analysis" This study investigates the impact of preparation for disaster relief on humanitarian response, where resilience mediates the relationship between preparation and response. The sample is composed of 1308 valid responses collected in China. The total indirect effect size is much stronger than the total direct effect size, which indicates the importance of resilience of humanitarian relief supply chain operations. A preliminary analysis of a moderated mediation model with uncertainty and respondents job characteristics as moderators presents interesting results. While the direct effect size of preparation on response and resilience is between small and medium, the effect size of resilience on response is very large, which indicates the importance of increased resilience. Small effects are observed for job characteristics on response, and the interaction effect of job characteristics and preparation on resilience. A very large simple effect is observed for uncertainty on resilience, even if the interactive effect seems much smaller. The theoretical and practical implications of the findings will be presented along the future research directions. 4) Dr. Bill Wang,, Kean University – Wenzhou Campus "Internet of Things (IoT)-Enabled accountability in source separation of household waste for a circular economy in China" Abstract Source separation is regarded as a best practice for sustainable waste management, which is essential for a transition to a circular economy to recover value from waste. However, its implementation in China has faced many difficulties which are primarily inherent in the public’s behaviors towards source separation. of household waste. Based on multiple cases of innovative use of the Internet of Things (IoT) technologies in China in recent years, this study establishes the concept of IoT-enabled accountability in household. waste source separation by utilizing the lens of accountability theory. Moreover, this research advances several propositions on the multiple dimensions of accountability mechanism to influence user behaviors. The study’s findings provide guidance to governments, technology providers and waste management. organizations on the use of IoT-based technological solutions for sustainable waste. management. It stimulates future research on the use of IoT technologies in managing people’s behavior. in a range of contexts beyond waste management alone. The study contributes to the growing literature. on smart waste management. | Ping Wang | James Madison University | Ping Wang | James Madison University | Xiaofei Zhao, Yinghong Dong, Ping Wang, Bill Wang |
34 | Frontiers in Multiple Hypothesis Testing | Multiple hypothesis testing is a long-standing yet top-trending research topic, generating numerous statistical theories and inference procedures for simultaneous hypothesis tests. These multiple testing procedures have been widely applied in diverse scientific domains such as biology, medicine, genetics, neuroscience, and finance. Although many of these approaches have become gold standards in many applied fields, cutting-edge research is continuously pushing this area to its frontiers. This invited session consists of four presentations that represent the recent advances of methods and theory in multiple hypothesis testing, including conformal inference, e-values, post-selection inference, and statistical inference on false discovery proportion. This session will be attractive to anyone who is interested in the state of the art in multiple hypothesis testing and related areas. | Yuan Jiang | Oregon State University | Yuan Jiang | Oregon State University | Wenguang Sun, Xianyang Zhang, Zijun Gao, Tao Yu |
35 | Go No Go decision making in clinical trails | Go no go decision is crucial in early clinical development. In this session, we will discuss go no go decision making using dual criterion or with multiple endpoints. Operational considerations in two-stage seamless transition design and optimization of study resources in phase II part of the phase II/III seamless design will also be shared. | Hongjie Deng | Boehringer-Ingelheim | Chen Lu | Boehringer Ingelheim | Chao Cheng, Lijiang Geng, Harry Hua, Bin Zhuo |
36 | High dimensional data analysis | This session will discuss new techniques and models that can handle high-dimensional data. | Juan Hu | DePaul University | Yichao Wu | University of Illinois Chicago | Yong Wang, Xu Liu, Jing Wang, Juan Hu |
37 | High dimensional hypothesis testing problems | Zhaojun Wang | School of Statistics and Data Science, Nankai University | Long Feng | School of Statistics and Data Science, Nankai University | Guanghui Wang, Lihua Peng, Lilun Du, Long Feng | |
38 | High Dimensional Statistical Inference and Its Applications | The topic of this session high dimensional statistical inference and its applications, which are including test and measure for partial mean; high-dimensional normality test; high-dimensional regression and model averaging, as well as high-dimensional mean vectors test. | Hengjian Cui | Capital Normal University, Beijing, PR.China | Hengjian Cui | Capital Normal University, Beijing, PR.China | Wei Zhong, Wenwen Guo, Guanpeng Wang, Lingyue Zhang |
39 | High-Dimensional Data Analysis and Variable Selection | High dimensional data analysis problems arise from diverse fields of scientific research and technological development. The ultra high dimensionality is a massive challenge for modern data analysis. There is an escalating demand for new efficient methods in dimension reduction and variable selection to enhance visualization, prediction, computation and interpretation of complex data. This session invites four insightful presentations, each provides latest theoretical and methodological developments in the area of high dimensional data analysis. The topics cover dimension reduction for complex tensor data, variable selection with control error rates, convergence rate of high-dimensional self-normalized Gaussian approximation, and semi-supervised learning in high dimensions. This session is designed to bring researchers from diverse subfields together and catalyze intellectual synergies across different areas. | Lan Gao | The University of Tennessee Knoxville | Lan Gao | The University of Tennessee Knoxville | Elynn Chen, Daoji Li, Jingkun Qiu, Yuqian Zhang |
40 | Inference for paired comparision data and clustering | When subjects are difficult to rank simultaneously by a judgment of one person, they are arranged to be repeatedly compared with one another in pairs. Paired comparison data also arise in situations in which there are natural win-loss results between two subjects without the presence of a judge. How to rank subjects and clustering has gained wide attention in machine learning and statistics literature. This session will bring us new methodologies and theories on paired companions and clustering. | Ting Yan | Central China Normal University | Ting Yan | Central China Normal University | Yuxin Chen, Ruijian Han, Nan Lu, Ting Yan |
41 | Innovations in Statistical Methodologies: From Functional Data Analysis to Spatial Modeling | This session aims to showcase recent advancements in statistical methodologies, bringing together experts and researchers to explore cutting-edge approaches in data analysis. The session will feature four presentations covering diverse topics, each contributing unique insights to the field. The content is accessible to both academics and industry professionals seeking practical insights and applications in data analysis. Attendees will gain insights into the latest developments in statistical methodologies, spanning functional data analysis, correction of measurement biases, shape mediation analysis, and spatial modeling with varying coefficients. The session aims to foster discussions and collaborations among participants, providing a platform for sharing knowledge and advancing the field. We believe that this session will contribute significantly to the overall theme of the meeting, providing attendees with a diverse and comprehensive overview of innovations in statistical methodologies. | Jing Wang | University of Illinois at Chicago | Jing Wang | University of Illinois at Chicago | Lijian Yang, Lan Xue, Jiangyan Wang, Shan Yu |
42 | Innovative design in oncology clinical trials | In recent years, innovative design has become increasingly important in clinical trials, especially in the highly expensive fields of oncology therapeutic area and special disease areas, such as rare diseases and pediatric medication. Regulatory agencies such as the US FDA/NMPA (China) have also drafted guidelines on adaptive design to guide pharmaceutical industry in their practices. This kind of design will bring more challenges in statistical methods than the traditional design. This session primarily invites a speaker with regulatory authority experience, one speaker from the academia, and one speaker from the pharmaceutical industry to discuss the application of innovative design in oncology clinical trials. | Dongmei (Ivy) Lu | Akeso Biopharma Company | Dongmei (Ivy) Lu | Akeso Biopharma Company | Xinxu Li, Fangrong Yan, Zhuozhao Zhan, Xianhua(Leslie) Meng |
43 | Innovative Machine Learning Approaches in Genomics and Genetics | This session brings together four leading researchers in the fields of genetics/genomics and bioinformatics, showcasing cutting-edge methods and applications in understanding complex biological mechanism. Dr. Yun Li from UNC Chapel Hill introduces "launch-dCas9," a machine learning-based framework for predicting gRNA impacts in CRISPR epigenome editing, enhancing the precision of genetic editing tools. Dr. Chong Wu from MD Anderson Cancer Center presents "PURE," an integrated approach for causal protein discovery that leverages both cis- and trans-acting elements, offering significant advancements in proteome-wide association studies (PWAS). From The University of Hong Kong, Dr. Yan Zhang discusses "TransTWAS," a novel multi-tissue transfer learning method for transcriptome-wide association studies (TWAS), addressing the challenge of limited sample sizes in specific tissues. Finally, Dr. Xiang Chen from St. Jude Children's Research Hospital introduces "Seq2Karyotype (S2K)," an innovative algorithm for deconvoluting heterogeneity of copy number alterations in cancer research using single-sample whole-genome sequencing. This session provides insights into the latest computational methods enhancing our understanding and treatment of complex diseases. | Cai Li | St. Jude Children's Research Hospital | Feifei Xiao | University of Florida | Yun Li, Chong Wu, Yan Dora Zhang, Xiang Chen |
44 | Lifetime Data Analysis | Recent advances of many aspects of lifetime data analysis will be discussed in this session. | Mei-Ling Ting Lee | University of Maryland | Xingqiu Zhao | Hong Kong.Polytechnic University | Jialiang Li, Shu-Hui Chang, Mei-Ling Ting Lee, Chung-Chou Chang |
45 | Machine Learning and Causal Inference Advances in Biostatistics | This session focuses on recent developments in biostatistics, specifically highlighting advancements in machine learning and causal inference. Presentations will cover topics such as the estimation of total mediation effects, contrastive dimension reduction methods, and biclustering methods with feature selection for omics data analysis. Additionally, insights into the reliability improvement of machine learning algorithms through random-effects split conformal interval prediction will be discussed. Join us for a comprehensive and scholarly exploration of these methodological advancements in the context of biostatistics. | Chong Wang | Iowa State University | Cindy Yu | Iowa State University | Chunlin Li, Yunhui Qi, Chong Wang, Peng Liu |
46 | Machine learning methods for multi-omics data analysis | Over the past decade, various sequencing technologies have revolutionized our understanding of complex biological processes, providing unprecedented resolution at the single-cell level and spatial localization information in tissues. By quantifying the genetic, epigenomic, and transcriptomic levels within individual cells and combining them with other screening techniques, these technologies have enabled the dissection of regulatory networks and cellular heterogeneity, offering crucial insights into disease mechanisms. As advancements in sequencing technologies continue, both the volume and complexity of data have significantly increased. The surge in genomic data has given rise to critical computational challenges in the design of genomic experiments and the efficient analysis of high-dimensional data, necessitating tailored and efficient methodologies, including advanced machine learning methods. Our session will center on developing machine learning methods for the analysis of multi-omics data, leveraging molecular and imaging features. We aim to provide novel biological insights within the field by harnessing the power of machine learning to uncover hidden patterns, identify biomarkers, and predict disease outcomes, thereby contributing to the advancement of precision medicine and personalized healthcare. | Ying Ma | Brown University | Ying Ma | Brown University | Fenghai Duan, Lin Lin, Xuegong Zhang, Shihua Zhang |
47 | Machine learning methods for survival data | This session consists of four speakers who are senior faculty members from universities in Canada, Hong Kong, and China and have established records of research in machine learning methods for survival data. They will present topics around tree-based methods, including Polya tree, Bayesian tree, and boosting decision tree, and a deep learning method for issues such as complex covariate effects, mediation analysis, and longitudinal covariates in survival data. | Yingwei Peng | Queen's University | Yingwei Peng | Queen's University | Liqun Diao, Yingwei Peng, Xinyuan Song, Xuejing Zhao |
48 | Moderate Deviation and Inference for Independent and Dependent Data | This session is concerned with recent developments in moderate deviation principle and statistical inference for independent data, time series and spatial processes (also called random fields) with various dependence structures and/or heavy-tailed distributions. In particular, the speakers will present their research on moderate deviations for i.i.d. data, martingales, linear random fields and modularity in network, and study the statistical inference of these moderate deviations. | Hailin Sang | University of Mississippi | Hailin Sang | University of Mississippi | Yongcheng Qi, Xiequan Fan, Guangyu Yang, Hailin Sang |
49 | Modern Statistical Methodologies for Analyzing Complicated Data | This session has four speakers, who will talk about the resampling approach for massive data with measurement error, semiparametric approaches for left-truncated and interval-censored survival data from the complicated clinical design, and the nonparametric approach for causal inference. | Yuanshan Wu | School of Statistics and Mathematics, Zhongnan University of Economics and Law | Yuanshan Wu | School of Statistics and Mathematics, Zhongnan University of Economics and Law | Jianwen Cai, Haibo Zhou, Mingqiu Wang, Jichang Yu |
50 | Modern statistical methods for time series data | The session consists of four experts in the time series area. They will present cutting-edge research topics on statistical methods for tail-dependent time series, nonparametric methods for long-range variance estimation, and new autoregressive models and simultaneous inferential procedures that can simultaneously characterize influence and temporal dependence in time series data. | Shujie Ma | UC-Riverside | Shujie Ma | UC-Riverside | Kin Wai Keith Chan, Ting Zhang, Weichi Wu, Jie Li |
51 | New Development in Causal Discovery and Biomarker Identification for Precision Medicine | In recent years, precision medicine has been the focus of both academia and industry. The rationale behind precision medicine is to tailor medical decisions and interventions to the specific characteristics (e.g., transcriptomic data) of each patient, rather than adopting a one-size-fits-all approach. Although this rationale has led to many important methodology developments in this field, challenges in precision medicine remain, especially how to efficiently use high-dimensional data for biomarker discovery and decision making. The goal of this invited session is to provide recent developments on how to address these challenges from the perspective of both methodological development and translational research. | Muxuan Liang | University of Florida | Muxuan Liang | University of Florida | , , , |
52 | New developments in health data analysis with applications to clinical trials | The session is about new developments in complex biomedical data analysis with applications to clinical trials. The speakers in the session are excellent experts in the research fields. The speaker will explore optimal treatment allocation policies that target distributional welfare. Depending on the choice of the quantile probability, this criterion can accommodate a policymaker who is either prudent or negligent. The speaker will introduce minimax optimal policies that are robust to model uncertainty. Interval-censored failure time data occur in many areas, including demographical studies, economic studies, medical studies and social sciences, and in different forms. In the session, the speaker will discuss variable selection for such data and present some recently developed tools. Other talks in this session include gene expression data analysis and sequential monitoring in covariate adaptive randomized clinical trials with nonparametric approaches, etc. All the talks in this proposed session have new applications in the applied research. | Yichuan Zhao | Georgia State University | Yichuan Zhao | Georgia State University | Yifan Cui, Tony Sun, Tiejun Tong, Hongjian Zhu |
53 | New developments in modeling high-dimensional and complex data | This session focuses on introducing the recent developments in modeling high-dimensional and complex data. - Conformal knockoff conditional independent test with double robustness - Normalized power prior Bayesian analysis: setup and computation - Enhanced Feature Screening in Ultrahigh Dimensional Semi-competing Risks Data via Sufficient Dimension Reduction - A new sufficient variable screening approach via characteristic functions | Wenbo Wu | The University of Texas at San Antonio | Wenbo Wu | The University of Texas at San Antonio | Baoying Yang, Keying Ye, Chenlu Ke, Pei Wang |
54 | #REF! | #REF! | #REF! | #REF! | #REF! | #REF! | Zhiguang Huo, Jinyuan Liu, Ruitao Lin, Muxuan Liang |
55 | New Developments of Statistical Methods in Model Selection and Bio-Medical Studies | The speakers (from Canada, USA and China) of this session will discuss new developments of statistical methods in model selection and bio-medical studies. Topics include: 1. Variable Selection after Multiple Imputation on Epidemiological Data with Missing Values. 2. A Constrained Minimum Criterion for Regression Model Selection. 3. Estimation of volume under the ROC surface with verification bias. 4. Distributed statistical inference for multi-source data. | Gengsheng Qin | Georgia State University, USA | Gengsheng Qin | Georgia State University, USA | Ye Shen, Min Tsao, Zhouping Li, Gengsheng Qin |
56 | New methods in nonparametric statistics with applications to time series | The session is about new methods in nonparametric statistics and new applications. The speakers in the session are excellent experts in the research fields. The speaker will propose new methods of interval estimation for the correlation coefficient using jackknife empirical likelihood, mean jackknife empirical likelihood and adjusted jackknife empirical likelihood. For better performance with small sample sizes, the speaker will also propose mean adjusted empirical likelihood. The speaker will study the test for serial or cross dependence in time series data that are subject to underreporting. The talk will introduce new test statistics, develop corresponding block bootstrap techniques, and establish their consistency. The methods are shown to be efficient by simulation and are used to identify key factors responsible for the spread of dengue fever and the occurrence of cardiovascular disease. In many scientific and technological fields, multivariate functional data is often repeatedly observed under different conditions within a given time period. It is of utmost importance to determine if the mean vector function for this type of data is consistently equal throughout the entire period. In the session, the speaker will introduce four novel global testing statistics that employ integrating and maximizing techniques to address this issue. Other talks in this session include oracle-efficient M-estimation for single-index models with a smooth simultaneous confidence band, etc. All the talks in this proposed session have new applications in the applied research. | Yichuan Zhao | Georgia State University | Yichuan Zhao | Georgia State University | Linlin Dai, Yingcun Xia, Li Cai, Zhiping Qiu |
57 | New Methods on Statistical Learning with Survival Data | The proposed session will present recent methodological developments that contribute novel analytical tools for probing treatment effect or other biomedical effects of interest on survival outcomes. The new methods deal with various realistic complications, including competing risks, aggregated data, confounding, and time-varying treatment/exposure. These methods improve over existing ones in terms of modeling flexibility, estimation efficiency, or robustness. This session will facilitate the dissemination of a set of state-of-art statistical tools and foster substantive applications in a variety of scientific context. | Limin Peng | Emory University | Yijuan Hu | Emory University | Yu Cheng, Huijuan Ma, Qi Zheng, Limin Peng |
58 | New nonparametric and dimension reduction methods for causal inference and clustering analysis | Over the past few decades, there have been significant achievements in the development of causal inference and clustering analysis (subgroup analysis) theory and methods. These methods have a wide range of applications in different fields, including epidemiology, biostatistics, statistics, computer science, and economics. This session consists of four experts in this area to present new nonparametric and dimension reduction methods for causal inference and clustering analysis. The new methods provide a flexible and powerful approach to analyzing complex data settings including spatial data, functional data, and heavy-tailed data. | Shujie Ma | University of California, Riverside | Shujie Ma | University of California, Riverside | Guanyu Hu, Wei Luo, Sijian Wang, Shujie Ma |
59 | New Technology and Quantitative Tools in Drug Development | Recently, as the new technology evolving, many applications have been realized in all processes of drug development. In this section, we introduce several new tools, such as R package or AI, which play substantial roles in quantitative science and faciliate drug development. | Fan Wu | Junshi Bioscience Inc. | Leslie Meng | Boehringer-ingelheim.com | Leslie Meng, Xiaocong Zhang, Xin Zhang, Bingying Xie |
60 | Novel Applications of Advanced Statistical Learning Methods to Biomedical Data | Biomedical data, derived from sequencing and clinical trials, are inherently complex and multi-dimensional. They often exhibit characteristics such as non-linearity, high dimensionality, and heterogeneity, posing significant challenges in analysis and interpretation. This session aims to delve into the application of novel statistical learning methods to navigate these complexities, offering innovative solutions for biomedical data analysis. | Jian; Li-Xuan Zou; Qin | Chongqing Medical University; Memorial Sloan Kettering Cancer Center | Jian Zou | Chongqing Medical University | Yupeng Cun, Judy (Hua) Zhong, Kevin (Zhi) He, Jian Zou |
61 | Novel Statistical Methods with Applications in Medical and Financial Studies | This session consists of four speakers. Their talks focus on recent developments in novel statistical modeling approaches for biomedical studies and financial data analysis. Topics include a Gaussian process joint model for analyzing mixed mode data, factor-augmented regression analysis, predictor screening and selection in the context of high-dimensional quantile regression, and a novel approach for analyzing financial data. | Xinyuan Song | The Chinese University of Hong Kong | Xiangnan Feng | Fudan University | Kai Kang, Deng Pan, Xuejun Jiang, Bing Wang |
62 | Public health practices under data science | We have four speakers with detail below, including speaker, institute, and email 1..Tang Li, St. Jude Graduate School of Biomedical Sciences, USA, li.tang@stjude.org, Unlocking the potential of statistical engineering in precision healthcare 2..Liu Catherine, The Hong Kong Polytechnic University, macliu@polyu.edu.hk Copula-enhanced CNN in joint prediction of refraction error and axial length based on ultra-widefield fundus images 3..Luo Rui, City University of Hong Kong, ruiluo@cityu.edu.hk Blockchain-Enabled Social Media Network for Mitigating Misinformation Spread 4..Pan Yinghao, UNC Charlotte, ypan8@uncc.edu Survival Bandits | Catherine Liu | The Hong Kong Polytechnic University | Catherine Liu | The Hong Kong Polytechnic University | Li Tang, Catherine Liu, Lorry Luo, Yinghao Pan |
63 | Real World Evidence: Under the Regulatory Microscope | In recent years, Real World Evidence (RWE) has been a hotspot in clinical and statistical research. Scientists and regulatory bodies, such as FDA and CDE, are interested in exploring the possibility of utilizing RWE in clinical trials. A few years ago, FDA initiated an RCT-Duplicate study with Harvard, Duke and Aetion, to investigate when, how and where RWE can "duplicate" the results from Randomized Clinical Trial. In 2023, FDA awarded four regulatory sciences projects, one of them being "Integrative Analysis of Randomized Clinical Trials with Real-World Data: Methods and Applications" with focus on hidden bias, to Prof Xiaofei Wang of Duke. Since 2019, Chinese government has set Hainan Island as the "RWE pilot zone", to bring in drugs and medical devices already approved overseas to China with RWE, without formal clinical trials. In this session, we are going to learn the RCT-Duplicate study in China, by Prof Feng Sun, of Beijing University; the FDA awarded RCT-RWE joint analysis study by Prof Xiaofei Wang, of Duke. In addition, two seasoned statisticians, Dr. Bo Jin will speak about statistical rigors in RWE analyses for regulatory consideration, while Dr. Sheng Feng will introduce the Hainan RWE pathway. RWE: how does it look like under regulatory microscope? | Sheng Feng | Parexel | Sheng Feng | Parexel | Xiaofei Wang, Feng Sun, Jie Chen, Sheng Feng |
64 | Recent Advances and Challenges in complex statistical modeling and Data Analysis | This session focuses on the recent developments in complex statistical modeling and the related applications. We provide a wide range of methodological and practical analyses based on statistical learning, like distribution-free prediction bands for clustered data, a power-robust test for global hypotheses in generalized models, and a correlation coefficient measure for blocking-selection models. | Yuehan Yang | Central University of Finance and Economics | Yuehan Yang | Central University of Finance and Economics | yanlin Tang, Yaowu Liu, Yuehan Yang, Siwei Xia |
65 | Recent advances in analysis of complex networks data | Networks are very common in a wide variety of fields, including social sciences, biological sciences, transportation systems, and power grids. Quantifying the influence of network features on edge formation is a key issue in network analysis. For this purpose, some new theories and new models have been proposed for the recent years, including semiparametric network models, latent space models and stochastic block models, etc. The speakers in this session will introduce their latest research findings in such field. | Hui Zhao | Zhongnan University of Economics and Law | Xiaobing Zhao | Zhejiang University of Finance and Economics | Jianwei Hu, Lianqiang Qu, Xiaobing Zhao, Hui Zhao |
66 | Recent advances in Bayesian modeling of spatially resolved transcriptomics data | Molecular profiling technologies such as genome sequencing have transformed life sciences. Innovative spatially resolved transcriptomics (SRT) techniques, which enable transcriptome measurement in high spatial resolution, have achieved significant breakthroughs in recent years. Due to the enormous potential to deepen our understanding of molecular mechanisms, these new SRT technologies are rapidly gaining attention, and a massive amount of gene expression data with spatial information has been generated. These emerging data will undoubtedly lead to an explosion of innovations in spatial statistics, an area fueled by breakthrough technologies in data collection and rooted in model-driven statistical research. To fully understand the diversity and potential impacts of spatial distributions of gene expression, customized statistical models are urgently needed. The molecular profile of SRT data is typically a high-dimensional count matrix. Analyzing such multivariate count data remains challenging because it is unnormalized, over-dispersed, and zero-inflated. Those characteristics make standard analytic methods less powerful and no longer applicable. Advanced Bayesian statistical inference has shown strength in analyzing genomics data (e.g., microarray and RNA-Seq) with a small sample size but a large number of features. The Bayesian approach naturally overcomes those challenges. We foresee that it will result in a new wave of innovation in statistics, enhancing our knowledge of biological mechanisms from spatial analysis. | Qiwei Li | The University of Texas at Dallas | Qiwei Li | The University of Texas at Dallas | Xiangyu Luo, Bencong Zhu, Qiwei Li, Qiwei Li |
67 | Recent advances in complex network | The proposed session will present recent developments for complex networks. The new models and methods will be introduced to deal with various realistic problems for different networks datasets, including directed and bipartite networks, co-occurrence networks, attributed networks, and networks with missing edges. These novel models and methods enhance existing methods and fill in some blank research areas. This session will facilitate the dissemination of a set of state-of-art methods and foster substantive applications in a variety of scientific context. | Jingli Wang | Nankai University | Jingli Wang | Nankai University | Bingyi Jing, Rui Pan, Pengsheng Ji, Yuan Zhang |
68 | Recent Advances in Data Integration and Structure Identification | Large scale data from a single or multiple sources are becoming increasing prevalent and are constantly evolving due to technology advancements. However, these data are subject to challenges such as high-dimensionality, data dimension mismatch, missing data, as well as censoring. New statistical methods to handle complex scenarios in data integration and structure identification are in great demand, hence motivating the proposal of this session. The primary objective of this session is to present statistical challenges, innovations, and applications pertaining to the modeling of complex data. Drs. Ding, Shu, Tang, and Yu, four renowned researchers who are at the forefront of this area, will present various new models, methods, and empirical studies related to (1) causal subgroup identification for survival outcome, (2) latent factor decomposition for high-dimensional data, (3) recommendation under missing data, and (4) classification under block-wise missing data. The speaker names, affiliations, and their presentation titles are listed below. | Lu Tang | University of Pittsburgh | Lu Tang | University of Pittsburgh | Ying Ding, Hai Shu, Xiwei Tang, Guan Yu |
69 | Recent Advances in Design of Experiments and Subsampling | This session includes four speeches on Recent Advances in Design of Experiments and Subsampling. | Min-Qian Liu | Nankai University | Jian-Feng Yang | Nankai University | Fasheng Sun, Yongdao Zhou, Jianbin Chen, Chunyan Wang |
70 | Recent advances in high dimensional complex data analysis | High dimensional data with complex structures has become increasing prevalent in many scientific fields nowadays and opened a new era of Big Data allowing an exponential increase of discoveries through statistical learning. Adapting to the growing data dimensionality and complexity, substantial developments have been made in the areas of dimension reduction, survival analysis, nonparametric analysis, among others. This session will highlight frontier research in high-dimensional complex data analysis such as dimension reduction in time series data, statistical learning of multi-sourced matrix-valued data, nonparametric models for complex discrete high-dimensional data and copula models for competing risks data. The audience will be exposed to key notions in high dimensional analysis, practical tools for statistical inference, and timely insights into various research areas. Therefore, the proposed session will appeal to a wide range of graduate students, researchers and practitioners who are interested in the mining of high dimensional complex data. This session will feature four speakers: 1. Dr. Yue Cui (YueCui@MissouriState.edu) is an Assistant Professor of Statistics in the Department of Mathematics at Missouri State University. Her research focuses on effect-size based nonparametric models for clustered data and high-dimensional data within Quality-of-Life measurements. She has papers and R packages published in the subject of informative and interpretable effect-size nonparametric models for complex data when distributional assumptions can not fulfill. She will present “Effect-size Based Nonparametric Methods for Complex High Dimensional Data”. 2. Dr. Antai Wang (aw224@njit.edu) is an Associate Professor of Statistics in the Department of Mathematical Science at New Jersey Institute of Technology. His main research interests include survival analysis and microarray data analysis, leading to publications in premier journals like Statistical Sinica and Biometrika. He will bring the talk titled in “The identifiability of copula models for competing risks data with exponentially distributed margins”. 3. Dr. Jiaying Weng (jweng@bentley.edu) is an Assistant Professor of Statistics in the Department of Mathematical Sciences at Bentley University. Her research areas encompass sufficient dimension reduction, sufficient variable selection, and nonparametric analysis. She has published papers in Journal of Nonparametric Statistics and other multidisciplinary areas. She will introduce a new dimension reduction method in her talk, namely, “Dimension reduction for nonlinear vector autoregressive models”, 4. Dr. Chenglong Ye (chenglong.ye@uky.edu) is an Assistant Professor in the Dr. Bing Zhang Department of Statistics at the University of Kentucky. His broad research interests include data integration, model combining, variable importance and statistical applications. He has publications in top-tier journals like Journal of American Statistical Association, Journal of Computational and Graphical Statistics, etc. He will bring a talk on “High-Dimensional Learning for Multi-Sourced Matrix-shaped Data”. | Yue Cui | Missouri State University | Chenlu Ke | Virginia Commonwealth University | Yue Cui, Antai Wang, Jiaying Weng, Chenglong Ye |
71 | Recent advances in high-dimensional statistical inference | In the era of big data, where information is amassed at an unprecedented rate and complexity, the importance of advanced statistical methods in unraveling patterns within high-dimensional datasets cannot be overstated. The exponential growth in data dimensionality continues to pose new challenges, demanding new, sophisticated analytical approaches, and statistical methods for high-dimensional data analysis play a pivotal role in the process. In this session, I invite speakers of various academic backgrounds and seniority who have recent breakthroughs in developing methods for high-dimensional data analysis in diverse settings. They include: (1) Dr. Yue Wang, who is an Assistant Professor of Biostatistics at the University of Colorado Anschutz Medical Campus, will present his work on direct estimation of differential Granger causality between two high-dimensional time series; (2) Dr. Xiang Zhan, currently an Associate Professor at the Beijing International Center for Mathematical Research and the Department of Biostatistics at Peking University, will present his work on nonparametric composition-on-composition regression analysis for high dimensional microbiome data; and (3) Dr. Songshan Yang, an Assistant Professor in the Institute of Statistics and Big Data at Renmin University of China, will present an asset splitting algorithm for ultrahigh dimensional portfolio selection and its theoretical property. Finally, I myself will present a graph-informed high-dimensional analysis framework and its applications in disease-associated gene set discovery and disease risk prediction. By organizing this session, I hope to gather researchers with expertise in high-dimensional statistics and start a discussion on promising future research directions in the field, with applications in various areas such as genetics/genomics, social science, finance, public health, and medicine. | Jin Jin | University of Pennsylvania | Jin Jin | University of Pennsylvania | Yue Wang, Xiang Zhan, Songshan Yang, Jin Jin |
72 | The Inference of Generalized Win Odds | It is a common practice to collect complex longitudinal and survival data arisen in clinical trial and observational studies. It is important to explore and infer the dependence and association between the dynamics of longitudinal process and clinical endpoint (event) so that the bias from simultaneous inference can be reduced and the efficiency of the estimates can be improved. Joint modeling and associated statistical methods for longitudinal and survival data have been an active area that has received much attention in statistical research. In this session, some challenging advances and issues of modeling approaches and statistical algorithms for joint analysis of complex longitudinal and survival data collected in health and biomedical fields will be addressed and further research directions will be discussed. This invited session will provide an opportunity to promote new development of joint modeling and associated statistical methods for longitudinal and survival data with applications in the fields and the session audiences will be benefited by this important statistical topic. | Yangxin Huang | University of South Florida, USA | Changyong Feng | University of Rochester, USA | Lang Wu, Changyong Feng, Jianghu Dong, Yi Xiong |
73 | Recent advances in modeling single-cell data | The session brings together researchers from different fields including statistics, biostatistics, and bioinformatics. The speakers will speak about their recent works on modeling single-cell data from association to causal analyses. | Yang Ni | Texas A&M University | Qiwei Li | UT Dallas | James Cai, Yuchao Jiang, Fangda Song, Yang Ni |
74 | Recent advances in modeling spatial transcriptomics data | Molecular profiling technologies such as genome sequencing have transformed life sciences. The development of innovative spatial transcriptomics techniques, which enable transcriptome measurement in high spatial resolution, has achieved significant breakthroughs in recent years. Due to the enormous potential to deepen our understanding of molecular mechanisms, these new technologies are rapidly gaining attention, and a massive amount of gene expression data with spatial information has been generated. These emerging data will undoubtedly lead to an explosion of innovations in computational methods. We foresee the new spatial data will result in a new wave of innovation in statistics and data science, deepening our knowledge of biological mechanisms from spatial analysis. | Qiwei Li | The University of Texas at Dallas | Yang Ni | Texas A&M University | Xiang Zhou, Mingyao Li, Guanghua Xiao, Jin Liu |
75 | Recent Advances in Modern Data Science | The session "Recent Advances in Modern Data Science" promises to be an enlightening exploration of cutting-edge developments in the field. This dynamic session will feature presentations from four distinguished speakers, each poised to share their insights, research findings, and expertise in various facets of modern data science. From innovative methodologies to groundbreaking applications, attendees can expect a comprehensive overview of the latest trends and advancements shaping the landscape of data science. This session provides a unique opportunity for participants to engage with leading experts, fostering a collaborative atmosphere that encourages knowledge exchange and dialogue. Join us for an intellectually stimulating experience that delves into the forefront of modern data science and its transformative impact on diverse industries. | Ping Ma | University of Georgia | Cheng Meng | Renmin University of China | Yuehua Cui, Huimin Cheng, Terry Ma, Jingyi Zhang |
76 | Recent Advances in Network Data Analysis | Network data are frequently used in various fields like genetics, sociology, finance, and econometrics. Networks consist of nodes and edges linking one node to another. How to characterize the generating mechanisms of networks is a key issue in network analysis. Many statistical models have been proposed to address this issue. This session will bring us recent methodologies, theories and techincal tools for network data analysis. | TING YAN | Central China Normal University | Binyan Jiang | The Hong Kong Polytechnic University | Binyan Jiang, Wanjie Wang, Will be updated later Will be updated later, Will be updated later Will be updated later |
77 | Recent Advances in Response Adaptive Randomization | This session will explore Response-Adaptive Randomization (RAR), a sophisticated approach in experimental design where allocation probabilities are dynamically adjusted. This method is crucial for reducing exposure to less effective treatments and improving the precision of our findings. Our agenda includes four cutting-edge studies in the field of RAR. Traditional theories in RAR are typically tied to specific parametric models, but we are now witnessing a growing recognition of RAR’s broader applicability. The first two presentations will explore the resilience of RAR beyond these traditional models. The initial presentation will discuss regression-based approaches within a model-independent framework, followed by a talk on the integration of semiparametric methods to enhance accuracy and statistical power. The significance of surrogate endpoints in clinical trials, especially for the rapid assessment of new treatments, cannot be overstated. The third segment of our session will introduce an innovative approach that incorporates these surrogate endpoints into both the design and analysis stages, providing empirical evidence of its effectiveness. Finally, addressing the variability in treatment effects is a common challenge in clinical research. Our concluding presentation will introduce a Bayesian-adaptive randomization technique, designed specifically for identifying diverse treatment effects, while also considering essential statistical factors. This session aims to provide a comprehensive overview of the latest developments in RAR, emphasizing both theoretical advancements and practical applications. | Yang Liu | Renmin University of China | Wei Ma | Renmin University of China | Wei Ma, Fuyi Tu, Jingya Gao, Zhongqiang Liu |
78 | Recent advances in statistical genetics and genomics | Genetic and genomic research has witnessed unprecedented growth, driven by technological advancements and interdisciplinary collaboration. This session on “Recent advances in statistical genetics and genomics” aims to provide an introduction of recent breakthroughs in statistical methodologies that underpin genetic and genomic analyses. Key topics include advanced methods for cell lineage reconstruction, identification and putative causal variants, and cell type deconvolution and annotation. These methods will demonstrate how latest statistical methods extract valuable insights from genetic and genomic data, to enable a deeper understanding of cellular heterogeneity within tissues and genetic factors contributing to complex traits and diseases. | Wei Vivian Li | University of California, Riverside | Wei Vivian Li | University of California, Riverside | Jiajun Zhang, Hao (Harry) Feng, Minghua Deng, Yi Yang |
79 | Recent Advances in Statistical Inference on Data Integration | Delve into the forefront of statistical inference with the session "Recent Advances in Statistical Inference on Data Integration." This dynamic exploration unveils innovative approaches reshaping how meaningful insights are extracted from integrated datasets. The session highlights recent developments, including singular propensity score techniques transforming the statistical landscape in data integration, semi-parametric quantile regression's power in combining diverse samples, and the transformative potential of group-sparse inductive matrix completion through transfer learning. Additionally, participants will gain insights into groundbreaking techniques challenging traditional assumptions in observational studies, expanding the scope of causal inference in integrated datasets. Tailored for statisticians, data scientists, and researchers, this engaging session offers a unique opportunity to navigate recent breakthroughs and discuss practical applications. | Zhengyuan Zhu | Iowa State University | Zhengyuan Zhu | Iowa State University | Guoliang Ma, Kosuke Morikawa, Hengfang Wang, Cindy Yu |
80 | Recent advances in statistical methods for analyzing two-phase data and genetic data | This session focuses on recent advances in statistical methods for analyzing complex biomedical data including two-phase data and genetic data, each with two speakers. With two-phase data, one speaker will focus on the inference of nonparametric variable importance and another speaker on sampling design for causal inference. With genetic data, one speaker will focus on statistical algorithm for cancer subtyping and the another speaker on statistical method for discovering driver gene, both through mutually exclusive genetic signatures. | Min Yuan | Anhui Medical University | Hong Zhang | University of Science and Technology of China | Min Yuan, Guorong Dai, Zheyu Zhang, Min Zeng |
81 | Recent Advances in Statistical Methods for Complex Data Analysis | This session brings together leading researchers in the field of statistics and data science to present their latest advancements in statistical methods for analyzing complex data. The focus is on innovative approaches to nonparametric estimation, regression models, and classification techniques, addressing challenges such as high-dimensional data, group structure identification, and dynamic classification. | Yue Niu | The University of Arizona | Yue Niu | The University of Arizona | Hao (Helen) Zhang, Zhengyuan Zhu, Yuan Ji, Ning Hao |
82 | Recent Advances in Statistical Methods for Complex Imaging Data | Imaging data is complex, requiring sophisticated statistical methods to be developed. This session features recent developments in this area, with a particular focus on addressing the complexities of modeling imaging data, including its integration with other types of data. Four excellent speakers will discuss various recent advances in bridging this gap. | Xi Luo | Univ of Texas Health Science Center at Houston | Xi Luo | Univ of Texas Health Science Center at Houston | R Todd Ogden, Robert Krafty, Yi Zhao, Zhengwu Zhang |
83 | Recent advances in statistical methods for complex time-to-event data | The goal of this invited session is to bring together four speakers who are experts and active researchers in time-to-event data analysis. They will present newly developed statistical methods for modeling time-to-event data, including high-dimensional confidence intervals for censored quantile regression, optimal treatment regimes in the presence of a cure fraction, a flexible varying coefficient model for survival prediction, and estimation and regression with sequentially truncated data. All of these presented methods are motivated by current real clinical or epidemiological studies and are rigorously studied through both theoretical justifications and comprehensive numerical studies. The four speakers will demonstrate to the audience that successfully addressing these analytical challenges is urgent and crucial for the efficient design of Alzheimer's disease studies, as well as for precision disease monitoring and prevention in cancer and cardiovascular diseases. The proposed session will be beneficial to both statisticians interested in methodology development in this area and investigators interested in applying new and sophisticated statistical methods to analyze real-world time-to-event data. This session consists of four speakers and a session chair with diverse backgrounds, including different countries/regions (United States, Hong Kong, and mainland China), different genders, and various years of experience. | Jing Qian | University of Massachusetts Amherst | Yubo Luo | Beijing Technology and Business University | Tony Sit, Cunjie Lin, Wen Li, Jing Qian |
84 | Recent Advances in Statistical Methods with Applications in Biomedical Research | This session consists of four well-established biostatisticians who are going to present a variety of novel statistical methods applied to biomedical research. The session has covered a few extremely popular research topics in biostatistics, including Bayesian models for meta-analysis, multimodal data integration, and centile chart methods which are popular for assessing measurements that vary by age. The session also comprises various biomedical applications. Dr. Qingxia Chen will present data from clinical trials used for meta-analysis; Dr. Ting Li will present the fusion of imaging, genetics, and clinical data; Dr. Tianying Wang will present rare disease data; Dr. Siyuan Ma will present microbiome data. The presenter structure of the session is also diverse. We have three female presenters and one male presenter; two senior faculty (full and associate professors) and two junior faculty (assistant professors). | Panpan Zhang | Vanderbilt University Medical Center | Ran Tao | Vanderbilt University Medical Center | Ran Tao, Ting Li, Tianying Wang, Siyuan Ma |
85 | Recent advances in statistical modeling and computation in applications with complex data structures | Statistical modeling and computation in applications are powerful tools that provide a formal framework for data-driven decision-making processes, resulting in their widespread usage for different inferential and predictive purposes. Furthermore, recent developments in these techniques from both Bayesian and frequentist perspectives offer novel insights, interpretation, and convenience through the development of efficient sampling algorithms and the incorporation of prior information in Bayesian inference. This is particularly well-suited to practical scenarios with complex data structures that commonly arise in biomedicine, clinical studies, economics, industrial engineering, and big data applications. The proposed session will illustrate complementary aspects of statistical modeling and computation with residual sampling techniques, covering established regression models to recent deep learning architectures and model averaging mechanisms. Dr. Zhuanzhuan Ma will discuss the shrinkage of effect size when external data are borrowed in clinical study settings. Dr. Min Wang will introduce a closed-form Bayes factor to test the equality of two sample mean vectors in high-dimensional settings. Dr. Liucang Wu will propose a novel residual subsampling algorithm for a skew-normal mode regression model with massive data. Finally, Dr. Linhan Ouyang will present a novel sparse regression Kriging procedure for the mean function that ensures prediction accuracy while using only a limited number of variables to capture the potential existing trend. Four expository talks and a subsequent panel discussion will encourage the audience to consider adopting recent advances in statistical modeling and computation in applications to analyze complicated data, investigate relevant factors, and leverage statistical learning models. Thus, we are confident that the topics of this invited session are highly related to the symposium theme “Innovation in Data and Statistical Science: Theory, Methodology, and Practice” and should be of particular interest to the 2024 ICSA China Conference audience. | Min Wang | The University of Texas at San Antonio | Liucang Wu | Kunming University of Science and Technology | Zhuanzhuan Ma, Liucang Wu, Min Wang, Linhan Ouyang |
86 | Recent advances in survival data modeling | This session consists of four speakers who are junior and senior faculty members from universities in Canada and China and have established records of research in areas of survival analysis. They will present topics around classical topics in survival analysis, including estimation methods for a flexible class of parametric and semiparametric survival models, subgroup analysis for clustered survival data, additive hazards cure models, and survival analysis with non-probability samples. | Yingwei Peng | Queen's University | Yingwei Peng | Queen's University | Wenqing He, Ye He, Hua Shen, Zhongwen Zhang |
87 | Recent advances in transfer learning and domain generalization | This session will focus on transfer learning, domain generalization, and related topics. Four speakers will discuss novel methodologies, statistical theory, and relevant applications in this field. | Sai Li | Renmin University of China | Huazhen Lin | Southwestern University of Finance and Economics | Yong He, Junlong Zhao, Mingyang Ren, Taoning Li |
88 | Recent advances on high-dimensional inference and generative models | This session focuses on some recent developments in data sciences and high dimensional statistics. | Yingying Fan | USC | Jinchi Lv | USC | Zhigang Bao, Fan Yang, Zemin Zheng, Yuting Wei |
89 | Recent development in change-point detection problem | Change-point detection is to detect distributional changes in a sequence of observations. It has important applications in many fields. The change-point detection problem has been well studied in the standard setting. However, modern data acquisition technologies have made possible the gathering of different types of data, which poses many challenges for traditional change-point detection procedures. In this invited session, four leading experts for the change-point detection problem will present their latest research on how to address those unique challenges and develop efficient change-point detection procedures in the non-standard settings. | Jun Li | University of California, Riverside | Jun Li | University of California, Riverside | Yajun Mei, Yanhong Wu, Hao Chen, Daren Change point inference in regression models |
90 | Recent development in statistical process monitoring | Statistical process monitoring applies statistical methods to the monitoring of a process in order to detect abnormal variations of the process. It has been successfully applied in various fields, including manufacturing, financial services, disease outbreak surveillance and network traffic monitoring. Recent advances in data acquisition technologies have made many different types of data available, which provides many unique opportunities for statistical process monitoring research. In this invited session, four leading experts in statistical process monitoring will present their latest research on how to develop efficient statistical process monitoring tools for some challenging situations in the real world. | Jun Li | University of California, Riverside | Jun Li | University of California, Riverside | Fugee Tsung, Dongdong Xiang, Wendong Li, Sven Knoth |
91 | Recent development in survival data and event history data analysis | Surival data and recurrent event data often occur in many fields including economics, medical studies and social science, and there has been a lot of work done on these data In the past few decades. As the development of information and measurement technology, the research about these data faces challenges of complex censoring mechanism, high-dimensionality, informative observation process and so on. To solve these new problems, some new theories and new methods have been proposed for the recent years, including variety of variable selection or screening methods, empirical likelihood methods, etc. The speakers in this session will introduce their latest research findings in such field. | Hui Zhao | Zhongnan University of Economics and Law | Ni Li | Hainan Normal University | Qingning Zhou, Huiqiong Li, Hong Wang, Ni Li |
92 | Recent development of joint modeling of longitudinal and survival data for population health research | Longitudinal data and survival data are commonly encountered data types in biostatistics research. In many applications, it is necessary to model these two types of data jointly to avoid bias, improve efficiency, or address the research question. In recent years, this field of research has demonstrated rapid growth toward incorporating more complicated data structures, such as multivariate or high-dimensional nonlinear longitudinal trajectories, or integrating with other statistical research areas such as risk prediction and causal inference. This session brings together four experts working at the forefront of this research field with their latest research presentations. | Liang Li | The University of Texas MD Anderson Cancer Center | Liang Li | The University of Texas MD Anderson Cancer Center | , Ruosha Li, Chixiang Chen, Liang Li |
93 | Recent developments for analyzing complex heterogeneous data | This session will showcase cutting-edge techniques, tools, and frameworks designed to extract meaningful insights from heterogeneous data. Discussions will revolve around novel algorithms, machine learning models, and statistical approaches specifically crafted to handle the intricacies of diverse data. Topics to be explored include but are not limited to: * hidden Markov models * semiparametric mixture regression * Penalized Robust Regression Estimator for High-dimensional Heterogeneous Data * Online Change Point Detection for Functional Data Attendees will gain valuable insights into the evolving landscape of data analysis, discovering innovative strategies to effectively extract knowledge and value from heterogeneous data sets. | Weixin Yao | University of California, Riverside | Weixin Yao | University of California, Riverside | Mian Huang, Yunlu Jiang, Hanbing Zhu, Weixin Yao |
94 | Recent Developments in Complex Time-to-Event Data Analysis | This session consists of four speakers. Their talks focus on recent developments in complex time-to-event data analysis. Topics include statistical inference for advanced survival models with length-biased data, ultrahigh-dimensional imaging covariates, current status data, and informative partly interval-censored data. | Xinyuan Song | The Chinese University of Hong Kong | Hong Wang | Central South University | Haijin He, Shuwei Li, Qi Yang, Jingjing Jiang |
95 | Recent developments in experimental designs | Experimental design is the process of planning studies and investigating efficient methods for collecting data for scientific investigation and is commonly used in industrial, agricultural, biological, pharmaceutical, manufacturing sciences, etc. The objective of an optimal experimental design is to provide interpretable and accurate inference at minimal costs. The purpose of this session is to present the current trends and developments in optimal designs. Topics include studying and comparing optimality criteria, selecting optimal designs from misspecified analysis of covariance models, and constructing designs and selecting models for mixture-process variable experiments when the number of variables is large. The applications of the optimal designs in industrial and manufacturing sciences will also be discussed. | Po Yang | University of Manitoba | Po Yang | University of Manitoba | Chang-Yun Lin, Xiaojian Xu, Po Yang, Xinwei Deng |
96 | Recent Developments in Longitudinal Data Analysis | This session focuses on presenting recent developments of statistical inference approaches and their applications to longitudinal data. | Xingqiu Zhao | The Hong Kong Polytechnic Universoty | Xingqiu Zhao | The Hong Kong Polytechnic Universoty | Juan Du, Kin Yau Wong, Shirong Deng, Dayu Sun |
97 | Recent developments in machine learning theory and application | Over the last few decades, we have witnessed rapid growth in machine learning. Machine learning algorithms have achieved great success in applications in many areas. In this session, I have invited four top experts in this field to discuss the recent developments in machine learning methods, theory, and applications. They will talk about several cutting-edge research topics in machine learning, including transfer learning for high-dimensional data, contrastive learning, robust estimation, and statistical learning theories in deep learning and applications. | Shujie Ma | University of California, Riverside | Shujie Ma | University of California, Riverside | Kun Chen, Cong Ma, Xuewei Cheng, Xinyu Zhang |
98 | Recent Developments in Network and matrix time series Analysis | In this section, we will introduce the recent development for network modelling. Specifically, we will give talks about network propagation model and discuss efficient estimation of SAR model for large-scale network data. Next, we will introduce modelling methods for matrix time series data. e.g., matrix quantile factor model or a bubble GARCH model. | Yingying Ma | Beihang University, School of Economics and Management | YIngying Ma | Beihang University, School of Economics and Management | Rongmao Zhang, Long Yu, Yingying Ma, Hong Chang |
99 | Recent Developments in Statistical Machine Learning | This session focuses on presenting recent developments of statistical machine learning approaches and their applications to complex data. | Xingqiu Zhao | The Hong Kong Polytechnic University | Xingqiu Zhao | The Hong Kong Polytechnic University | Xiaodong Yan, Xiangbin Hu, Qiang Wu, Yu Chen |
100 | Recent Developments of Bayesian Methods | In this session, four speakers will discuss recent developments of Bayesian methods in various applications. | Guanyu Hu | University of Texas Health Science Center at Houston | Guanyu Hu | University of Texas Health Science Center at Houston | Huiyan Sang, Yanxun Xu, Zhenke Wu, Yuexi Wang |
101 | Recent developments of statistical and computational methodologies for genomic data | With the advancement of technology and resource in genomic data collection and sharing, the variety and amount of genomic data have been rapidly growing, which provides an unprecedented opportunity for revolutionizing the landscape of medical and biological inquiry. Yet, this influx of diverse and vast genomic datasets presents tremendous analytical hurdles such as pleiotropic variants, intricate 3D genomic structures, biobank-scale sample size, high-dimensional data, subtle variant effects on complex diseases, and non-linear genetic-phenotypic relationships. To surmount those obstacles, the field of statistical and computational methodologies for genomic data has undergone a meteoric evolution. Our session convenes a panel of four experts in statistical genomics to present their new research works in the field. The spectrum of presentations will span from heritability estimation to multi-trait Genome-Wide Association Studies (GWAS), genetic admixture estimation, and the deconvolution of bulk RNAseq data deconvolution. This session is designed to attract individuals interested in either the development or application of statistical and computational methods for genomic data. It unveils some of the latest advancements, fostering a space for collaboration and innovation among enthusiasts of this rapidly progressing field. | Chenxi Li | Michigan State University | Chenxi Li | Michigan State University | Shili Lin, Kai Wang, Baolin Wu, Shaoyu Li |
102 | Recent statistical methodology advances in genomics and bioinformatics | Recent biotechnology and sequencing technique advances enable novel ‘omics data collections in a high-throughput fashion. In these data analysis, it is imperative to develop novel and suitable statistical methods tailored for such data. Methodology research in analyzing high-dimensional and high-throughput genomics and bioinformatics data, especially those data generated from novel biotechnology platforms, is drawing great attention recently. In this session, four speakers will share their recently developed statistical and analytical methods in this area. This session focuses on innovative approaches to analyze novel data types in genome, spatial transcriptome, epitranscriptome, and microbiome. | Hao Feng | Case Western Reserve University | Hao Feng | Case Western Reserve University | Wei Vivian Li, Zhenxing Guo, Yushu Shi, Zhonghua Liu |
103 | Recent statistical methods and applications in precision medicine | Precision medicine is a new approach for disease prevention and treatment that takes into account individual differences in people’s genes, environments, and lifestyles. Statistical methods promote the evidence-based decision making in achieving the precision medicine. In this section, four invited speakers will introduce their innovative work with applications to survival data, matrix-variate data and omics data. | Xiaoqing Pan | Shanghai Normal University | Xiaoqing Pan | Shanghai Normal University | Yuexin Fang, Zengchao Xu, Xiaoqing Pan, Pengyuan Liu |
104 | Recent studies on machine learning: from theory to application | Machine learning has made significant progress and continues to be a rapidly evolving field in recent years. Though having demonstrated exceptional performance in many tasks, machine learning methods have several limitations which researchers are actively working to overcome, for example, lack of interpretability and limited data efficiency. In this session, we would talk about theories that provide insights into how models arrive at their predictions, as well as applications in diverse domains to extract insights from large volumes of data and aid in predicting outcomes. | Xiao Wang | Qingdao University | Xiao Wang | Qingdao University | Xiaohui Yang, Rong Yin, Yang Yu, Baichen Yu |
105 | Regression Analysis of Failure Time Data with Complex Structures | Failure time data occur in many areas, especiallu in clinical trials or medical studies and their analysis has continuously been a hot topic partly because of the constant occurrences of new problems such as different complex data structures. In this session, we will discuss several such problems, including interval censoring and missing covariates. | Chunjie Wang | Changchun University of Technology | Shuying Wang | Changchun University of Technology | Chunjing Li, Xinrui Liu, Rong Liu, Wenshen Wang |
106 | Reinforcement Learning and Personalized Medicine | This section focus on developing new reinforcement learning tools for dynamic decision-making problems. Many application problems have been viewed as reinforcement learning, such as personalized medicine and pricing, while various statistical challenges are involved including unobserved confounders and quantifying uncertainties. This proposed session includes several very active researchers in this field. | Ruoqing Zhu | University of Illinois Urbana Champaign | Yifan Cui | Zhejiang University | Zhengling Qi, Will Wei Sun, Ruoqing Zhu, Tao Shen |
107 | Reinforcement Learning from A Statistical Perspective | Reinforcement learning (RL) is garnering a flurry of interest in recent years. A central objective of RL is to search for a policy — based on a sequential collection of noisy data samples — that approximately maximizes cumulative rewards, without direct access to a precise description of the underlying environment. In contemporary applications, it is increasingly more common to encounter environments with prohibitively large state and action spaces. The explosion of model dimensionality exacerbates the challenge of achieving efficient RL in sample-starved applications, where data collection is expensive, time-consuming, or even high-stakes (such as clinical trials, online advertisements, autonomous systems, to name just a few). The challenge is further compounded by the daunting nonconvexity issues intrinsic to natural RL formulations, presenting a major roadblock towards understanding the efficacy of the RL algorithms in use. Consequently, understanding and improving the statistical efficiency of RL algorithms inevitably lie at the core of cutting-edge RL research and are the key enabler for future advances. Despite decades-long research, however, the statistical underpinnings of RL remain far from mature, especially when it comes to finite-sample analyses that are of crucial operational values in practice. In truth, a large body of RL literature focused on asymptotic analyses (where the number of samples tends to infinity with the model size held fixed), while non-asymptotic analyses for various RL algorithms remained largely unavailable until fairly recently. The inadequacy of asymptotic analyses for the sample-starved regime (where the growth of model complexity outpaces that of the sample size) is readily illuminated from lessons learned from the investigation of high-dimensional statistics: the large-sample asymptotics systematically fail to capture the statistical and algorithmic bottlenecks in the sample-limited regime. To enable faithful decision making in contemporary RL applications, it is in imminent need to obtain a more refined picture of the trade-off between sample complexity and statistical accuracy, and to design efficient algorithms that provably achieve the optimal trade-off. Encouragingly, there has been growing effort from the statistics community aimed at addressing these emerging challenges from diverse angles. This session aims to bring together leading experts in statistical RL to share their most recent findings, and to brainstorm important future directions that can enhance RL efficiency from the statistical perspectives. We have now lined up four excellent speakers working on the frontier of statistical RL. The tentative titles of these speakers are as follows. | Yuxin Chen | University of Pennsylvania | Yuxin Chen | University of Pennsylvania | Ruohan Zhan, Gen Li, Zhimei Ren, Zhengyuan Zhou |
108 | Robust Learning Inference for Data Science | This invited session will bring together four experts presenting some cutting-edge developments on robust learning inference for data science applications. It will involve a diverse set of invited speakers and research topics. | Jinchi Lv | University of Southern California | Yingying Fan | University of Southern California | Xiaohui Chen, Jinchi Lv, Zeng Li, Jingyuan Liu |
109 | Semiparametric panel data models' theory and applications | This session focuses on new developed theory and technique in semiparametric panel data models, as well as their applications. | Chaohua Dong | Zhongnan University of Economics and Law | Tingting Cheng | Nankai University | Fei Liu, Huanjun Zhu, Yayi Yan, Tingting Cheng |
110 | Semi-Supervised and Multi-Task Learning | The semi-supervised and multi-task learning paradigms have become effective ways to use unlabeled data and enhance model performance on a variety of tasks. Four speakers will provide an extensive exploration of the theory and applications of the cutting-edge approaches, and also discuss possible future directions. | Pengkun Yang | Tsinghua University | Huazhen Lin | Southwestern University of Finance and Economics | Xiaojun Mao, Yuan Cao, Han Zhao, Anru Zhang |
111 | Some Recent Advances in Survival Analysis | This session will focus on some of the new methods recently developed for regression analysis of survival analysis. In particular, the session will cover the topics of quantile regression, the analysis of interval-censored data and the deep learning method for survival data. | Mingyue Du | Jilin University | Hui Zhao | Zhongnan University of Economics and Law | Chunjie Wang, Peijie Wang, Shuying Wang, Mingyue Du |
112 | Spatial Data and Semiparametrics | Spatial data and non/semi-parametric modeling have been widely studied in the fields of statistics and econometrics. This Session will introduce recently developed methodologies in the areas of nonparametric tests and spatial autoregressive (SAR) models. The discussion on SAR models will include topics such as nonlinear transformation, matrix-valued data and semiparametric time varying coefficient models. | Jin Liu | Nankai University | Jin Liu | Nankai University | Zixin Yang, Kai Yang, Yimeng Ren, Jin Liu |
113 | Spatial data modeling and applications in biomedical data analysis | As technology continues to advance, spatial data has been as a key player in elucidating complex relationships within biomedical datasets. Notably, the rapidly advancing field of spatially resolved genomics presents an unprecedented opportunity to collect both molecular information and spatial context. This session directs its focus on the spatial data modeling and its power in analyzing and understanding the spatial dimensions of biomedical data. The cutting-edge techniques, state-of-the art software, applications, and challenges will be discussed. | Guoshuai Cai | University of Florida | Guoshuai Cai | University of Florida | Bo Cai, Xiaofei Zhang, Qianqian Song, Guoshuai Cai |
114 | Special considerations in oncology drug development | Oncology studies have many unique perspectives in drug development compared with non-oncology drugs. In this session, we invited speakers to talk about special considerations in oncology covering the whole drug develop cycle, including dose finding stage, dose optimization and pitoval stage. Specifically, the backfilling stragety in dose-escalation studies, efficient dose optimization designs, OS analysis and potential bias for regional treatment effect in phase III multi-regional clincial trials will be extensively discussed. | Yuan Geng | Daiichi Sankyo (China) Holdings Co., LTD. | Yuan Geng | Daiichi Sankyo (China) Holdings Co., LTD. | Michael Lee, Xi Huang, Zhiji Tang, Chengyuan Song |
115 | Statistical Advances in Modern Biomedical Applications: Robustness, Privacy, and Heterogeneity | In this session, four invited talks will explore exceptional statistical methods for biomedical data analysis. The initial presentation introduces a groundbreaking two-head neural network with a maximum rank correlation loss, highlighting its efficiency and robustness in handling diverse learning tasks, including classification and regression. The second talk focuses on innovative approaches for spatially variable gene identification using summary statistics in single-cell RNA sequencing studies. The third talk introduces an association test procedure based on summary statistics to identify pleiotropic effects with genetic variants. The final talk discusses the development of a robust, high-dimensional graphical model using gamma divergence to address dataset heterogeneity. These studies derive insights from various biomedical datasets, revealing compelling statistical findings. | Sheng Fu | National Cancer Institute at USA | Sheng Fu | National Cancer Institute at USA | Huichao Li, Han Yan, Deliang Bu, Sanguo Zhang |
116 | Statistical Analysis of Complex Data | This session is about some new developments of methods for complex data, such as two-stage clinical trials with covariate-dependent randomization, gut microbiome data, multiple heterogeneous studies with a cure fraction, and functional data. | Liuquan Sun | Academy of Mathematics and Systems Science, Chinese Academy of Sciences | Lianqiang Qu | Central China Normal University | Wei Zhang, Peng Ye, Bo Han, Xin Chen |
117 | Statistical and computational advances for complex biological data analysis | The proposed session addresses the growing need for advanced statistical and computational methods to analyze complex biological data. The speakers, each experts in their respective fields, will present state-of-the-art techniques and applications that have the potential to transform the way we approach and understand complex biological phenomena. This session aligns with the conference theme by fostering collaboration between statistical science and biological research, providing a platform for researchers to exchange ideas and explore new frontiers in interdisciplinary research. Specifically, Prof. J. Wang's talk will delve into the integration of multi-omics data within the context of 3D chromatin structure. The presentation will shed light on complex disease mechanisms, providing valuable insights into the intricate relationships between various biological factors. Prof. H. Wang's presentation will focus on the development of empirical likelihood-based semiparametric inference techniques for longitudinal data. The application to GAW 18 genetic data will showcase the practical significance of these methods in real-world scenarios. Prof. Li will discuss novel approaches for multi-marker genetic association and interaction tests, leveraging the accelerated failure time model. This promises to enhance our understanding of the genetic underpinnings of complex traits and diseases. Prof. He's talk will center around a novel ensemble feature selection approach, with a specific application in repertoire sequencing data. This innovative method will aid in extracting relevant features from high-dimensional biological datasets. The topics covered by the speakers represent the forefront of research, offering attendees valuable insights and fostering collaborative discussions. We are confident that this session will contribute significantly to the success of 2024 ICSA China conference and appeal to a broad audience interested in the intersection of statistics and biology. | Yuehua Cui | Michigan State University | Yuehua Cui | Michigan State University | Jianrong Wang, Honglang Wang, Chenxi Li, Tao He |
118 | Statistical Challenges in Medical Device Clinical Trials | The Food and Drug Administration (FDA) has established a three-tiered classification system for medical devices. Device classification depends on the intended use of the device and also upon indications for use. • Class I devices are low-risk devices o Examples include bandages, handheld surgical instruments, and nonelectric wheelchairs. • Class II devices are intermediate-risk devices o Examples include computed tomography (CT) scanners or infusion pumps for intravenous medications. • Class III devices are high-risk devices that are very important to health or sustaining life. o Examples include pacemakers, heart valves, and deep-brain stimulators. To get FDA clearance to market, certain type of premarketing submission/application is required according to the class to which the device is assigned. General Controls are the baseline requirements of the Food, Drug and Cosmetic (FD&C) Act that apply to all medical devices, Class I, II, and III. For Class I or II device, if it is not exempt, a 510k will be required for marketing. For Class III devices, a premarket approval application (PMA) will be required unless the device is a preamendments device (on the market prior to the passage of the medical device amendments in 1976, or substantially equivalent to such a device) and PMA's have not been called for. In that case, a 510k will be the route to market. The clinical trial for Class III device has similar challenges as Pharmaceutical or Biotechnology clinical trial, such as complexity of the study protocol, identification of suitable subjects etc. However, it has its unique characteristics. In general, medical device trials involves high cost, has longer follow-up time (normally 5 years or 10 years). Blinded, randomized, controlled trials (RCTs) is the gold standard design for drug trials but they rare within medical devices trials, as device trials are often very difficult to blind. Device studies often include highly diverse endpoints. In this session, the statistical analysis challenges in medical device clinical trials will be presented. It will also be sharing some statistical considerations in clinical trial design and to introduce an innovative study design for medical device trials to address some of the challenges. | Xingyu Gao | Edwards Lifesciences | Xingyu Gao | Edwards Lifesciences | Yu Shu, Yanglu Zhao, Charlie Ahn, Gaohong Dong |
119 | Statistical considerations for rare disease and pediatric clinical trials with challenges in study design | Though randomized controlled trial is the golden standard of clinical trials, we may encounter many challenges in the are of rare disease and pediatric. We may need to consider the impact of dosing change, to leverage external data when a controlled arm is practically and ethically impossible. And in most scenarios, we are conducting global wise clinical trial incoporating data of patients from multiple regions and we need to show the consistency among regions with small sample size of patients within each region. This session will cover the most commonly issues we may encounter and discuss the statistical considerations when addressing them. Some practical cases will be generalized and shown as case examples. | Hao Zhang | Sanofi, Inc. | Hao Zhang | Sanofi, Inc. | Wenruo Hu, Weiya Xu, Erin Feng, Grace Lin |
120 | Statistical Considerations of Time-to-Event Analysis in Modern Oncology Clinical Trials | Time-to-event(TTE) analysis in oncology clinical trials is crucial and requires a comprehensive understanding of statistical methodologies to ensure accurate interpretation and robust conclusions. This session delves into key statistical considerations that play a pivotal role in the analysis of time-to-event outcomes, with a focus on their application in the dynamic landscape of oncology trials. The challenges of censoring rules in the analysis of progression-free survival (PFS) as well as event-free-survival (EFS) will be explored in this session. Strategies for minimizing bias and maximizing the efficiency of TTE analyses will be discussed, providing attendees with practical tools for optimizing study design and data interpretation. Another import factor that may impact the Time-to-Event outcome is the handling of covariates in survival analysis. We will unravel the intricacies of covariate adjustment techniques, illuminating their role in mitigating confounding factors and elevating the accuracy of TTE measurements. Also, this session will delve into the temporal aspects of oncology studies, particularly the optimal length of imaging evaluations. Timely and precise assessments are crucial for capturing meaningful clinical endpoints. statistical approaches will be discussed to determine the most effective imaging evaluation intervals, balancing the need for frequent evaluations with practical considerations and resource constraints. In summary, this session aims to equip researchers and statisticians with a deepened understanding of statistical considerations in time-to-event analyses within the context of modern oncology clinical trials. | Tao Zhang | Qilu Pharmaceutical | Tao Zhang | Qilu Pharmaceutical | Vivian Yin, Julie Cong, Zhihao Cheng, Yuqi Wang |
121 | Statistical Frontiers: Exploring Novel Techniques in Modeling, Testing and Beyond | Embark on a journey through the forefront of statistical research with a session that delves into several pivotal topics. Uncover insights into variable screening using trees and random forests, delve into separable metric spaces with a novel two-sample test, unravel outlier detection techniques for functional data, and navigate fairness with Bayes-optimal classifiers under group equity principles. This comprehensive session provides a spotlight on innovative methodologies and insights, enriching the ever-evolving landscape of statistical advancements. | Fuke Wu | Huazhong University of Science and Technology | Hui Jiang | Huazhong University of Science and Technology | Zhibo Cai, Bu Zhou, Jia Guo, Xianli Zeng |
122 | Statistical Genetics: Using Statistical Tools to Elucidate the Genetic Basis of Complex Traits | In recent years, the opening of multiple large-scale biological sample databases has provided unprecedented opportunities for studying the genetic mechanisms of complex diseases. However, analyzing large-scale biological sample databases faces severe data analysis and computing challenges. Large biological sample databases usually have the following characteristics: multi-ethnic (leading to high data heterogeneity), multi-phenotype (complex correlations between different phenotypes), multi-omics (data is of high dimensionality and noisy) and large data volume (data storage is difficult and calculation is slow). In this session, researchers developed computationally efficient statistical models suitable for large biological sample databases to explore the pathogenic mechanisms of complex diseases and predict disease risks. | Xiaoxuan Xia | School of Life Sciences, Tianjin University | Xiaoxuan Xia | School of Life Sciences, Tianjin University | Mingxuan Cai, Nayang Shan, He Xu, Xiaoxuan Xia |
123 | Statistical inference with lean assumption | Statistical inference with lean assumptions represents a methodological approach in data analysis that prioritizes simplicity and parsimony in the underlying assumptions made during the inferential process. Unlike traditional statistical methods that may rely on complex and stringent assumptions, lean assumption inference seeks to minimize unnecessary assumptions to enhance the robustness and generalizability of results. This approach acknowledges the inherent uncertainties in real-world data and aims to provide reliable conclusions without burdening the analysis with unnecessary complexities. By embracing a lean assumption framework, statisticians and researchers prioritize practicality and applicability, making their analyses more accessible and interpretable while maintaining a high level of statistical rigor. This methodology is particularly valuable in situations where data may not conform perfectly to classical assumptions, allowing for more flexible and adaptive statistical modeling without compromising the validity of the inferences drawn. | Zhigen Zhao | Temple University | Kai Zhang | University of Carolina at Chapel Hill | Wenbo Wu, Kai Zhang, Xin Xing, Zhigen Zhao |
124 | #REF! | #REF! | #REF! | #REF! | #REF! | #REF! | #REF! |
125 | Statistical Learning from a Modern Perspective | Statistical paradigms for machine learning applications are certainly not in short supply. Over the past few decades, statistical thinking and reasoning have been a major driving force towards interpretable, trustworthy, and reproducible machine learning. In turn, the broad array of machine learning topics, in conjunction with the accompanying (and ever-changing) resource constraints, has constantly served as a key source of motivation for the advancement of modern statistical theory. From classical large-sample asymptotics to more recent high-dimensional data analysis, the foundation of statistical learning has been steadily evolving in response to the challenges over time. The need for an expanded statistical toolbox continues --- and in fact, intensifies --- in the contemporary big data era. While technological breakthroughs facilitate an unprecedented scale of data collection, the growth of the model complexity (i.e., the number of parameters employed to represent the model) usually even outpaces the increase of available data samples. As a core solution paradigm in high-dimensional statistics, one might postulate that the unknown parameters of interest are well embedded in a low-complexity model, and proceed by fitting such a reduced-dimensional model. Despite its enormous success, this paradigm remains highly inadequate in addressing many key challenges that emerge --- and become increasingly more pervasive --- in modern machine learning settings. A prominent example concerns the advent of deep learning. Focusing on idealistic and tractable models that still capture modern attributes, statisticians have identified new settings and phenomena that, if thoroughly studied, might help explain the unreasonable effectiveness of modern machine learning and reconcile it with classical statistics insights. This session aims to bring together leading experts in statistical machine learning to share their recent research findings, and visions of important future directions that can enhance the theoretical understanding of machine learning from the statistical perspectives. | Yuxin Chen | University of Pennsylvania | Yuxin Chen | University of Pennsylvania | Yao Xie, Changxiao Cai, Yuejie Chi, Wenpin Tang |
126 | Statistical learning with multiple data sources | The session consists of 4 talks, all related to statistical learning with multiple data sources. More specifically, the first talk is on the keyword extraction for textual data, stemming from various sources such as online product reviews and scholarly publications on scientific discoveries. The second talk is on distributed learning over multiple Electronic health records (EHRs) databases without sharing patient-level data. The third talk is on the functional clustering analysis of prevalence trends simultaneously for a large number of diseases. The last talk is on the individualized and adaptive transfer learning through a new statistical strategy via subset-average estimating equation of multi-task learning. | Tiejun Tong | Hong Kong Baptist University | Xiaochen Zhang | Beijing Normal University, Zhuhai | Xinlei Wang, Qi Long, Chenjin Ma, Xiaochen Zhang |
127 | Statistical Methods and Applications for Healthy Ageing | As of the end of 2021, the elderly population aged 60 and above in China was 267.36 million, accounting for 18.9% of the total population; the population aged 65 and above was 200.56 million, accounting for 14.2% of the total population in China. We will present a few topics on statistical methods and applications for prmoting healthy ageing and long term care in the elderly. | Xiaojun Wang | Renmin University of China | Tao Sun | Renmin University of China | Tao Sun, Yueming Jin, Huiping Zheng, Yunlong Li |
128 | Statistical Methods for Network Data | In this session, four speakers will discuss recent developments for network data. | Guanyu Hu | The University of Texas Health Science Center at Houston | Xuening Zhu | Fudan University | Ji Zhu, Junhui Wang, Emma Zhang, Tianxi Li |
129 | statistical methods in causal inference and decision making | This session will explore recent advancements in statistical and machine learning methods for causal inference and decision-making. The key areas of focus include causal graph learning, multi-agent reinforcement learning, statistical inference under unmeasured confounding, and distributional learning. | Jiayi Wang | University of Texas at Dallas | Jiayi Wang | University of Texas at Dallas | Hengrui Cai, Wenzhuo Zhou, Yunan Wu, Yan Zhong |
130 | Statistical Methods of Survival Data | This session is about some new developments of methods for survival data, such as streaming survival data, dyadic link formations in directed networks, recurrent event data with non-negligible event duration, and imaging response data. | Liuquan Sun | Academy of Mathematics and Systems Science, Chinese Academy of Sciences | Zhang Wei | Academy of Mathematics and Systems Science, Chinese Academy of Sciences | Dongxiao Han, Lianqiang Qu, Xiaowei Sun, Wenliang Pan |
131 | Statistical modeling for complex data | In applications we often encounter complex data with different structures. In this session we introduce some new methods for modeling such kinds of data which put great challenges to researchers. Our methodology will contribute to econometrics, data science, and statistics, and stimulus others to continue working in related areas and apply our methods to solve real problems. | Jiancheng Jiang | University of North Carolina at Charlotte | Maozai Tian | Renmin University of China | Shaoxin Hong, Youxi Luo, Fei Chen, Jiancheng Jiang |
132 | Statistical modeling of functional data and related applications in biomedicine | This session focuses on the statistical modeling methods of high-dimensional functional data in biomedicine and their application in the biomedical field. This session will include theoretical researches on statistical methods presented by two PhDs and applied researches by the other two PhDs in the field of clinical trial. | Min Yuan | Anhui Medical University | Min Yuan | Anhui Medical University | Zhanfeng Wang, Chengqing Wu, Jiuzhou Miao, Haolun Ding |
133 | Statistics in Biosciences | Statistics in Biosciences (SIBS), the journal of ICSA, is published three times a year in print and electronic form. It aims at development and application of statistical methods and their interface with other quantitative methods, such as computational and mathematical methods, in biological and life science, health science, and biopharmaceutical and biotechnological science. This session includes four excellent speakers from the journal editorial board to discuss several hot topics in statistical research. | (Tony) Jianguo Sun | UNIVERSITY OF MISSOURI-COLUMBIA | (Tony) Jianguo Sun | University of Missouri | Joan Hu, Yanqing Sun, Hulin Wu, Yi Li |
134 | Strategic Approaches and Methodological Innovations in Oncology Trials: From Early Phase Design Challenged to Novel Adaptive Strategies | In the clinical development of oncology in the pharmaceutical industry, we are increasingly pursuing faster and more efficient clinical research. In order to accelerate clinical development of oncology, on one hand it is essential to optimize strategic approaches among the numerous choices of development strategies. On the other hand, it is helpful to utilize innovative designs, such as innovative early phase designs like phase I/II trial, and adaptive late phase designs include phase II/III trials. In this session, each speaker explores a distinct facet, including options and challenges in early-phase oncology design, China dose confirmation in FiH Oncology study, a phase II/III adaptive design, and a novel phase I/II design. We hope to explore how these strategic approaches and methodological innovations are pivotal in advancing oncology trials. | Jun Zhang | Astellas | Jun Zhang | Astellas | Jun Zhang, Wentian Guo, Lusha Wang, Zilu Zhan |