MSCS Capstone Conference

Spring 2019 Schedule


28th-29th March 


Thursday, March 28th

Time

Room 241

Room 243

Room 245

Room 205

Room 254

9:00

Rose Havener

Days and Months and Years, Oh My!  Incorporating Goal Setting into Web-Based Calendars

Esteban Lemus Wirtz

Surviving Death Row: The Effect of Race in Time to Execution

Graham Low

Major League Baseball Hall of Fame Ballot Longevity 

Thomas Shogren

Major League Baseball Hall of Fame Ballot Longevity

Sean Cheng

Finding Patterns in Pitching Data: A Data Visualization Software

9:40

Malini Sharma

Robopup: Innate Behavioral Programming and Robotic Applications

Sebastian Coll

I will survive: Assessing the success of Spotify’s streaming chart-toppers

Marie Julstrom

Sonnetbot

Bridget Reilly

Magical Assistant reveals Fibonacci Magic

Sam Hollenbach

Creating a Pitch Tracking System for the Macalester Baseball Team

10:10

Aman Rai

The Equalizer: Can They Tie It Up? 

Sojin Oh

It's-a-Me, Artificial Intelligence: Training AI to play Super Mario Bros

Jeremy Moss

What type of dog is that? - Dog Breed Recognition on Mobile Phones

Ji Sue Song

Waiting for Approval, a survival analysis of U.S. patents

Jonathan Scott

Epidemics on Jamaica (EoJ)

10:40

Chan Wang

Does gender play a role in faculty promotion and retention?

Andrew Galletti

A Framework for Song Stem Generation

Roland Munsil

Drive Like a Roomba: A "Blind" AI Learns to Race

Rane Knecht Bice

Patently unfair?: Survival analysis gives a glimpse of trends in US patent process

Milo Beyene

Building a Personalized News Feed

11:30

Lu Xian

Capturing Dynamics of Time-varying Systems with Topology

Andy Hayes

How Long Are Art Exhibits on Display in the MIA?

Khin Thiri Kyaw

Is the Minneapolis Institute of Art (MIA)’s collection diverse? 

Emma Clink

Survival of the Hippest: What Keeps You on the Charts?

Taneeya Rele

Hit or Miss? Exploring Product Management Trade-Offs in Mobile Game Development.

12:00

Haimeng Zhang

The Cocktail Party Problem: Voice Separation Using ICA Algorithms

Colton Wenig

Staying on top: How long till they equalize?

Wes Summers

Graphic Model Design and Refinement

Preeta Raghunathan

Ukraine Specific Risk Factors for Mental Health Disorders

Will Kann

Securing a Firebase App: The Danger of Default Rules

12:30

Zhiyu Lin

When and Why Does the U.S. Supreme Court Overrule a Precedent?

Everett Hommes

Banking on Nostalgia? A Study on the Longevity of Broadway Productions

Karen Perez Sarmiento

Baby pictures of the universe: Signal processing on the sphere

Andy Hayes

 Bot or Not? A Network analysis of bot accounts on Twitter

Trevor Zapiecki

Phishing Our Phriends: How susceptible are students to phishing attacks?

1:20

Haimeng Zhang

Rendering Ocean Waves in OpenGL

Risa Shirai

Connecting the Dots: Using Computer Vision to Design Better 5G Cellular Networks

Ojashvi Rautela

Spherical Harmonics of the Earth

Lawson Busch

Green Canvas

Trung Nguyen

Drawing models from bones

1:50

Julia Romare

#nailfashion, #thefriendlywash, and #saygrace: A tour of #chicagogram

Giang Duong

A compact data structure for MASSIVE streaming data

Spencer Grant

Better than the Game Itself: Infinite Background Terrain Generation in Computer Graphics

Ben Lewis

Bot or Not? Twitter-Bot Identification Using Cosine Similarity

Charlotte Fowler

Effects of traumatic experiences on mental health in Ukraine

2:20

Paige Pfeiffer

Hearing Colors: Relationships between Albums and Color

Tiffany Fan

Efficient Algorithms for Approximating Matrix Functions

Ojashvi Rautela

Ray Tracing – Capturing the path of Light

Xinyu Yang

A Computer Assistant to Write Cover Letters

Trung Nguyen

Building voters’ preferences

3:00

Giang Duong

Colorizing black-and-white photos in one click

Gloria Odoemelam

Saving the World Takes Time and Money

Eric Frost

Signed, Scanned, Delivered: A Comparative Analysis of Sign Language Recognition Systems

Luke Brown

Integrating Instruction in Game Design

Xinwei Huang

How Long Does A War Last?

3:30

Nikhita Jain

How much is a life worth?

Joe Fish

Why Evictions Happen: A Survival Analysis on San Francisco Homes

Duc Le

Facial Recognition (almost) in your pocket

Jiaxuan Huang

What makes a hit song: An analysis of songs' survival time on the Billboard Top 100 Chart

Friday, March 29th

Room 241

Room 243

Room 245

Room 205

Room 254

9:00

Jonathan Scott

I am a computer who understands users. AMA

Tianyou Li

Busted! Fake News Classifier

Fouad El Hamdouni

Can you spot it? The implementation of SMS spam messages detector using Machine Learning Techniques

Xue Rui

Global Peacekeeper - A Survival Analysis on War Duration

Erle Lei

Save or Change?: How different features affect the lifespan of a smartphone

9:40

Maitrayee Deka

Order in Chaos: Using Computational Topology to Study Biological Aggregations

Tianyou Li

How To Cover Your Tracks Using Signal Processing

Brighten Jelke

Virtual Chimes: User Interface Design for VR Environments

Alice Qingyu Zhu

Popularity versus Similarity: Simulating Network Growth

Xuchen Wang

Which Flight to Pick? Factors of Flight Cancellations and Delays

10:10

Anja Beth Swoap

"Shall I Compare Thee To Thy Nearest Neighbors?": Combining Semantic and Phonetic Relevance for Poem Generation

Eric Huang

Single, Parallel, Graphical: Solving All Pairs Shortest Path through Parallel Processing

Emerald Thole

The Gravity of Game Design

Emmanuel Diaz

Minecraft AI

Zachary Skluzacek

Is MIA Biased? How Our Favorite Art Museum Plays Favorites

10:50

Sarah Pujol

Maybe She's Born With It, Maybe It's The Shading Algorithm

Kristijan Peev

Survival Analysis in Capital Punishment: Death Row Inmates’ Sojourn time in Texas Prisons

Kelli Mandel

The Craft of Software Design: How to Give Your Code a Makeover

Ansel Colby

Music Materialized: Graphically Enhancing the DJ Experience

Zhen Tan

Are you changing your cell phone or changing the world?

11:20

Maitrayee Deka

Deal or No Deal: Examining the Effects of Ransom Payment on Hostage Durations

Matt Yang

Playing with Fire: A Survival Analysis of Wildfire Size and Duration

Andrew Cui

Building Better Poker AI Through Opponent Modelling

Kavya Shetty

“God, I Hope I Get It! I Hope I Get It!: The Impact of the Tony Awards on Broadway Show Longevity”

Katie Jolly

Digitizing, Districting, and Data: Evaluating Political Gerrymandering in Ohio

12:00

Dona Pantova

First Moment of Quadratic L-functions in Function Fields

Kayla Beckham

Building 3D Virtual Chimes: Interactive Feedback in Virtual Reality

Zachary Skluzacek

Fighting Twitter Clutter

Haroon Malik

Timer of the Equalizer : Survival Analysis of the leading team in La Liga

Xuchen Wang

Neural Network Techniques on Spam Text Detection

12:30

Amanda Doan

Supercomputing for the Broke College Student

Usman Hasan

Fire and Fury: A Survival Analysis of Wildfire Cause and Duration

Maggie Hohenstein

Citation "Queer"ies

Troy Lyu

You're Fired: On Soccer Head Coach Tenure Lengths

Ruiqi (Annie) Wang

Waiting in the Terminal: An Analysis of Flight Delays

1:10

Ruotong Wang

Is Sociology missing something? Limitations of traditional network data collection

Gabriel Brown

Pac-Sphere: Pac-Man on a sphere

Andy Han

Theatrobo: Exploring Robot Theater with Shakespeare

Blake Jones

Lose Yourself: Slim Shady does Combinatorics

Mark Coretsopoulos

Modeling Realistic Human Movement

1:40

Chen Dun

No need to Trust anyone: secure online voting system

Sam Armon

Changing Coordinates in Permutation Groups

Carl Salvino

Avoiding the B-word: a Survival Analysis of Corporate Defaults

Courtney Overland

Simulated Jenga: Using Unity to Create a Video Game

Amanda Doan

Syllable Separation and Consonant Classification with Energy Transforms

2:20

Haolin Chen

Elena Smith

Bot Eye vs Human Eye: Understanding the Brain’s Visual System through Modeling V1 Striate Cells

Tim Lipman

Bulls, Bears, and Black Swans, oh My!

Ariana Pooley

Predicting Music Genre Classifications: Music Signals and Machine Learning Models

Jeremy Andrew Krovitz

Assignment Planning for the Visual Learner


Days and Months and Years, Oh My! Incorporating Goal Setting into Web-Based Calendars

Student:        Rose Havener

Major:                CS

Faculty Adv:        Lauren Milne

Didn’t follow through with your New Year’s resolution? G-Cal may not actually be the best way to organize your life -- it allows for scheduling, but is not very good for planning how to achieve goals. In this talk, I will describe an ongoing project to build a new and improved short-term and long-term task visualization tool. With this tool, goals can be visualized to better understand progress over time so you can stay on track.


Surviving Death Row: The Effect of Race in Time to Execution

Student:        Esteban Lemus Wirtz

Major:                MATH/AMS        

Faculty Adv:        Leslie Myint

Criminal justice systems aren’t always just. In contemporary America, one of the most notorious biases in criminal justice is racial bias. How do such biases affect the time to execution of death row inmates?  In this presentation, I will use survival analysis methods to explore how a death row inmate’s race and victim’s race affect survival time on death row.


Major League Baseball Hall of Fame Ballot Longevity

Student:        Graham Low

Major:                MATH/AMS

Faculty Adv:        Lori Ziegelmeier

Being inducted into the Major League Baseball Hall of Fame is a complicated process with no precise methodology for selection.  Because of this there is a large element of subjectivity for players becoming members of the Hall of Fame.  Using survival analysis methods, a player's chance of induction into the hall of fame will be determined given each individuals various career statistics.


Major League Baseball Hall of Fame Ballot Longevity

Student:        Thomas Shogren

Major:                MATH/AMS

Faculty Adv:        Lori Ziegelmeier

What statistics make a player more likely to be inducted into the Major League Baseball Hall of Fame? There are four milestones that are considered to guarantee entry into the Hall of Fame, but is that actually the case? Using survival analysis, we looked at various statistics and milestones to see if they really increase the chances of induction. Finally, we look at the 2019 Hall of Fame ballot and compare the results to expectations.


Finding Patterns in Pitching Data: A Data Visualization Software

Student:        Sean Cheng

Major:                CS

Faculty Adv:        Shilad Sen

Baseball pitchers throw thousands of pitches over their careers; having their pitches tracked helps them analyze and identify their strengths and weaknesses. As opposed to the charting pitches on pen and paper, we developed BullpenTracker, a website that allows pitchers to easily and quickly chart each of their pitches on a computer, and visualize the data immediately.


Robopup: Innate Behavioral Programming and Robotic Applications

Student:        Malini Sharma

Major:                CS

Faculty Adv:        Susan Fox

The 21st century has seen robotic pets move from a futuristic dream into a budding reality, with inventions such as the Paro seal. These inventions bring forth new complications for robotics and artificial intelligence as robots must deal with accurately modeling behavior and correctly interpreting information about the world around them. In my project, I explore both of these areas through the use of a Lego robot.


I will survive: Assessing the success of Spotify’s streaming chart-toppers

Student:        Sebastian Coll

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

What types of songs do well in the modern American music streaming environment? What musical traits and attributes are more likely to succeed? Hip hop, trap music, and R&B are at their peak right now, we’ll see how long these genres survive on the billboard. Using data gathered from Spotify’s weekly streaming charts and Spotify’s computer-generated music genome, we’ll take a look at what makes or breaks a chart-topper and what influences its survival.


Sonnetbot

Student:        Marie Julstrom

Major:                CS

Faculty Adv:        Lian Duan

The drab, humiliating, ready kiss / and her suburban, crowded, great gain knot / regarding their dream per our grim, lost bliss / from a alert, slim, lean weight plus his pot." -Sonnetbot, 2019. Come see how this bot generates nonsensical Shakespearean sonnets!


Magical Assistant reveals Fibonacci Magic

Student:        Bridget Reilly

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

How many different ways are there to place n dominos (vertically or horizontally) on a 2 by n board? The answer: the nth Fibonacci number! How about placing dominoes on a 3 by n board? This is a much harder question and that we will solve using generating functions, which are a powerful machine that allows you to answer otherwise complex problems in a simple way.


Creating a Pitch Tracking System for the Macalester Baseball Team

Student:        Sam Hollenbach

Major:                CS

Faculty Adv:        Shilad Sen

To leverage the power of statistics, competitive baseball teams keep diligent records of every pitch thrown by each pitcher on the team. However, many teams, including the Macalester Baseball team, still use antiquated pitch tracking systems, like pen and paper charts. Bullpen Tracker is a mobile app and web interface for collecting raw pitching data, with tools for advanced statistical analysis that was impossible with previous methods.


The Equalizer: Can They Tie It Up?

Student:        Aman Rai

Major:                MATH/AMS

Faculty Adv:        Joe Benson

When playing soccer on the world stage, every minute counts. Can we use survival analysis to predict the time taken by a team to equalize the score after conceding the first goal? I will at look at home field advantage and other factors affecting the survival rate of a lead in the LaLiga to determine what really matters when you are on the clock.


It's-a-Me, Artificial Intelligence: Training AI to play Super Mario Bros

Student:        Sojin Oh

Major:                CS

Faculty Adv:        Getiria Onsongo

Super Mario Bros game is a simple game where you have to jump over the obstacles and avoid bumping into enemies. However, imagine teaching a robot how to play this game when it has no idea whether jumping over the enemies or running into them is a better move. How should it learn the rules? In this talk, I’ll introduce how I trained AI agent to play Super Mario Bros game using reinforcement learning.


What type of dog is that? - Dog Breed Recognition on Mobile Phones

Student:        Jeremy Moss

Major:                CS

Faculty Adv:        Lian Duan

Have you ever seen a dog and been unable to tell which breed it is? What if I told you there was an app that could do this for you? Come see how transfer learning can be used in mobile neural networks to classify dog breeds on a mobile device.


Waiting for Approval, a survival analysis of U.S. patents

Student:        Ji Sue Song

Major:                MATH/AMS

Faculty Adv:        Will Mitchell

Patents and their approval process are more interesting than you think. There are many aspects of a patent that determine how long it takes for their application to be approved, and they are not as obvious as they seem. This research uses USPTO patent data and both parametric and non-parametric modelling to explore some of these factors, and which are more important than others.


Epidemics on Jamaica (EoJ)

Student:        Jonathan Scott

Major:                MATH/AMS

Faculty Adv:        Shilad Sen

The Chikungunya virus (CHIKV) continues to wreak havoc in tropical countries. Hence it is imperative to be able to model the spread of this disease in order to prepare for the future. In this project I use the Epidemic on Networks Python library, Gephi and tools from Network Science to model the spread of a virus through Jamaica.


 Does gender play a role in faculty promotion and retention?

Student:        Chan Wang

Major:                MATH/AMS

Faculty Adv:        David Shuman

Fair timing of faculty promotion and retention is pivotal to the performance of research institutions. Are male and female faculty promoted at the same rate? After how many years in the tenure track are faculty typically promoted to associate professors and full professors? When do professors leave the tenure track? My research will answer these questions using multiple survival models.


A Framework for Song Stem Generation

Student:        Andrew Galletti

Major:                CS

Faculty Adv:        Getiria Onsongo

Ever writing a song and hit melodic or harmonic writers' block? My project is a framework for the generation of different harmonic and melodic pairs in sequence; it provides little snippets for when you just can't find the next novel chord or note in the melody. I have employed different genetic algorithms and associated probability distributions to generate these sonic samples.  optogenetics to induce spikes in a system has drastically improved the quality of recordings and can be done in vivo. I will talk about a new statistical model that can jointly infer spiking activity and neural connectivity from such data.


Drive Like a Roomba: A "Blind" AI Learns to Race

Student:        Roland Munsil

Major:                CS

Faculty Adv:        Lian Duan

Can an AI learn to race with no way to see the course it's racing on? Using a simple machine learning technique, I created a racing game AI that learns using only its own location and speed.


Patently unfair?: Survival analysis gives a glimpse of trends in US patent process

Student:        Rane Knecht Bice

Major:                MATH/ AMS

Faculty Adv:        Tom Halverson

Facebook has applied for thousands of patents for technology that monitors us in new ways--like using facial recognition to gauge reactions to posts--and seeks to predict life events like when our friends will die.  Pharmaceutical patents affect the availability of medications. The patent process profoundly affects us. So what do we know about it?  Do corporations control it? Whose applications get fast-tracked? Using survival analysis to analyze length to patent approval provides some answers.


Building a Personalized News Feed

Student:        Milo Beyene

Major:                CS

Faculty Adv:        Shilad Sen

Recommendation algorithms influence not only the kind of information we see online, but how we see the world. While this may seem harmless for movies on Netflix, the implications of influence in news article recommendation may be more sinister. Two other Macalester students and I built our own personalized news feed in an attempt to better understand, apply and compare news recommender systems.


Capturing Dynamics of Time-varying Systems with Topology

Student:        Lu Xian

Major:                MATH/AMS

Faculty Adv:        Lori Ziegelmeier

While the dynamic behavior of a time-varying system, for example, a moving school of fish or flock of birds, can be difficult to capture, topological tools offer new insights into data analysis and interpretation. One such method is a crocker plot which displays the topological information of a system at all times simultaneously in a 2-dimensional image. A crocker video, as an extension of the crocker plot, shows persistent structural features with nice stability properties.


How Long Are Art Exhibits on Display in the MIA?

Student:        Andy Hayes

Major:                MATH/AMS

Faculty Adv:        Lisa Lendway

This project took a deep dive into a dataset covering exhibit durations at the Minneapolis Institute of Art (MIA). We looked at a variety of factors having influence on the amount of time that art was on display, including country, medium, department, and more! Sit in on this talk to find out if the curators at MIA have inherent bias in the art they choose to display!


Is the Minneapolis Institute of Art (MIA)’s collection diverse?I

Student:        Khin Thiri Kyaw

Major:                CS

Faculty Adv:        Lian Duan

The MIA has one of the most wide-ranging art collections in the country, containing approximately 89,000 artworks from six continents and spanning over 20,000 years. But, is it as diverse as they say? Come to this talk to learn about the diversity of MIA’s collection with epic visualizations of metadata corresponding to the date, location, and medium of objects in the museum.


Survival of the Hippest: What Keeps You on the Charts?

Student:        Emma Clink

Major:                MATH/AMS

Faculty Adv:        Kristin Heysse

Throughout the years we’ve seen countless musicians achieve fame or fade away after one song. This makes you wonder— what causes some songs to become timeless classics but others to be forgotten? Through survival analysis we found that the energy, danceability, and more can help a song hold a spot on the charts for longer, thus increasing the chance of lasting popularity. Come find out how long your favorite songs would survive the charts


Hit or Miss? Exploring Product Management Trade-Offs in Mobile Game Development.

Student:        Taneeya Rele

Major:                CS

Faculty Adv:        Paul Cantrell

The 8 months of developing Flingshot, a student-made, cross-platform and reflex-based mobile game was full of trade-offs. Some were a hit, some were a complete miss…

This talk walks through the development history of Flingshot in hindsight, and applies the principles of product management to its development process. It explores the importance of the product manager role in game development because product management is like a game. The only difference is that in games you get a second chance to start afresh.


The Cocktail Party Problem: Voice Separation Using ICA Algorithms

Student:        Haimeng Zhang

Major:                MATH/AMS

Faculty Adv:        Lori Ziegelmeier

Imagine you are at a party with roughly 11 people speaking at the same time. You want to keep track of what each person is talking about, so you place 11 recorders in the room. How would this help us? This is a typical Blind Source Separation (BSS) problem. In this talk, I will present Independent Component Analysis (ICA) algorithms and show how to use them to extract individual voices from audio mixtures.


Staying on top: how long till they equalize?

Student:        Colton Wenig

Major:                MATH/AMS

Faculty Adv:        Joe Benson

Have you ever wondered what factors play a role in the equalization of a soccer match? Does home field advantage really make a difference? It this talk I will analyze these questions and more using survival analysis of the 2017/18 Professional soccer league in Spain.


 Graphic Model Design and Refinement

Student:        Wes Summers

Major:                CS

Faculty Adv:        Lian Duan

This talk focuses on the process of creating a virtual model, going through the initial rudimentary steps to showing how the model is gradually refined to bring it closer to reality, complete with an examination of how to apply textures to increase realism, as well as looking at the tools used to create and improve the model.


 Ukraine Specific Risk Factors for Mental Health Disorders

Student:        Preeta Raghunathan

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

Intermittent Explosive Disorder (IED) is an impulse control disorder characterized by violent physical or verbal outbursts that are disproportionate to the sensory trigger. Ukraine, located in eastern Europe has a long history with conflict and war due to its history with the Soviet Union. As a result, there is a high incidence of substance abuse, anxiety disorders, and depression. These factors  are known to be linked to the onset of IED, making Ukraine an ideal candidate for studying IED risk factors.


Securing a Firebase App: The Danger of Default Rules

Student:        Will Kann

Major:                CS

Faculty Adv:        Paul Cantrell

Firebase is a popular cloud database. In 2018, over 3,000 Firebase apps leaked sensitive user data to the public. Why? Because in many cases, the developers failed to write any security rules, allowing anyone read/write access to their database. I demonstrate the exploitability of an unsecured application by stealing data from Matchbook, an unreleased social network app. I then explain how developers can avoid leaking their data by writing specific security rules and performing regression tests.


When and Why Does the U.S. Supreme Court Overrule a Precedent?

Student:        Zhiyu Lin

Major:                MATH/AMS

Faculty Adv:        Lori Ziegelmeier

The Supreme Court has made many controversial decisions throughout history. Recently, it legalized same sex marriage, upheld Trump’s immigration ban, and blocked regulations on abortion clinics. How were these decisions made? Why during certain times? How do they relate to the historical and contemporary political environments? This project analyzes the Court decision data in the past 80 years using survival regression models and derives some insights on the judiciary’s role as a constrained policy maker.


Banking on Nostalgia? A Study on the Longevity of Broadway Productions

Student:        Everett Hommes

Major:                MATH/AMS

Faculty Adv:        Brianna Heggeseth

Most broadway shows close before their tenure even exceeds 3 months, but a few have enjoyed years of unremitting success. Why do certain productions last so much longer than others? Why does broadway keep producing repeat productions of old shows such as “Carousel” and “West Side Story”? Is broadway losing its originality over a desire to appease nostalgia? Come see these questions answered and more as techniques in Survival Analysis intersect the Fine Arts.


Baby pictures of the universe: Signal processing on the sphere

Student:        Karen Perez Sarmiento

Major:                MATH/AMS

Faculty Adv:        Leslie Myint

The data in scientific disciplines like astrophysics, geology, and geography is often collected on the globe or in the sky, thus making it spherical. How do we adapt conventional data analysis techniques for data that varies in one dimension to accommodate spherical 3D signals? In this talk, I will cover some background on spherical harmonic analysis, which helps us understand oscillatory features of the data. As an application, I will show how this tool helps us understand the history of the universe, starting from its baby picture.


 Bot or Not? A Network analysis of bot accounts on Twitter

Student:        Andy Hayes

Major:                CS

Faculty Adv:        Lisa Lendway

Are you on Twitter? If so, chances are you have encountered a bot at some point. Of the 300 million twitter accounts it is estimated that 40 million are fake. These fake accounts follow real users all the time and this poses the question: If bots follow real people, do they follow each other? If so, what can we learn from the resulting bot networks? Come to this presentation to find out!


Phishing Our Phriends: How susceptible are students to phishing attacks?

Student:        Trevor Zapiecki

Major:                CS

Faculty Adv:        Paul Cantrell

Phishing is a scamming technique where emails are sent pretending to be from a trusted source, and try to steal sensitive information from the receivers. As Macalester students studying computer science, we like to believe that we are fairly safe and won’t fall for scams. Is that really the case? In this talk, I will explain how we conducted an experiment in which we attempted to phish our classmates, without actually stealing any information.


Rendering Ocean Waves in OpenGL

Student:        Haimeng Zhang

Major:                CS

Faculty Adv:        Lori Ziegelmeier

Rendering realistic water remains a challenging and important aspect in computer graphics and animation. The Disney film Moana has its scenes set in the ocean and is highly praised for its ocean rendering effects. In this talk, I will explore how to produce realistic rendering of ocean surface, mimicking the reflection and refraction properties to achieve a realistic look of water surface.


Connecting the Dots: Using Computer Vision to Design Better 5G Cellular Networks

Student:        Risa Shirai

Major:                CS

Faculty Adv:        Susan Fox

Upcoming 5G cellular technology requires accurate modeling of tree density in the network path. I examined computer vision techniques that could be used to identify trees and their corresponding densities from aerial images. This might be the solution to the challenges faced by the telecommunications industry today. I experimented with convolutional neural networks, which are a form of deep learning where a network learns image features, given a large training sample.


Spherical Harmonics of the Earth

Student:        Ojashvi Rautela

Major:                MATH/AMS

Faculty Adv:        Ian Whitehead

Spherical harmonics are used to study data on a sphere, in this case, the Earth! Earth has a dramatic topography with its deep ocean basins and high mountain ranges. In geosciences, mathematical principles are used to understand the internal dynamics of the Earth. However, when we model the earth, we cannot ignore the structural difference between mountains and oceans. Here, we will explore how spherical harmonics capture this variability in Earth topography.


Green Canvas

Student:        Lawson Busch

Major:                CS

Faculty Adv:        Will Mitchell

Video game graphics evolved over the past decade, with the driving force being a desire to appear more realistic. That meant more physics in graphics. Green Canvas discusses and implements some of these methods in one of the most challenging graphics problems: grass blades.


Drawing models from bones

Student:        Trung Nguyen

Major:                CS

Faculty Adv:        Andrew Beveridge

If you have watched The Lord of the Rings (you should have), have you ever wondered how Gollum was animated? In a process called skinning, CGI magicians turn sensors into flesh and blood and actions into movie sequences. For my project, I study and implement a simple skinning algorithm widely used in industry, called linear blend skinning. Come find out the what goes into the toolboxes of Pixar and Disney animators!


 #nailfashion, #thefriendlywash, and #saygrace: A tour of #chicagogram.

Student:        Julia Romare

Major:                CS

Faculty Adv:        Lauren Milne

Can we learn common features, themes, and activities of a geographical area based on Instagram hashtags? In this talk, I will discuss how we use hashtags to convey something about our activities and online identities. We also use neighborhood names as hashtags, which gives a geographical area its own online persona. Join me for an exploration of content and geographical aspects of Chicago Instagram posts.


A compact data structure for MASSIVE streaming data

Student:        Giang Duong

Major:                MATH/AMS

Faculty Adv:        Leslie Myint

Imagine you have to analyze a never-ending data stream, think Google searches, to find the top trending searches for an ad client. Unfortunately, your hardware has very limited storage. To solve this problem, this project explores Count-Min Sketch, a super fast, highly memory-efficient probabilistic data structure that approximates statistical queries of massive frequency data without storing all incoming data. Come learn about this powerful tool: the theoretical framework, implementation and application in weak password detection.


Better than the Game Itself: Infinite Background Terrain Generation in Computer Graphics

Student:        Spencer Grant

Major:                CS

Faculty Adv:        Joe Benson

Picture yourself on a beautiful summer day while taking a hike through the woods. Are you simply focused on getting to the destination without taking in your surroundings? Me neither. So why should an electronic game focus entirely on the gameplay? Here I will discuss a process of generating a terrain that is infinite, randomized, and realistic that can enhance the aesthetic of the gaming experience using techniques including lighting models and memory optimization.


Bot or Not? Twitter-Bot Identification Using Cosine Similarity

Student:        Ben Lewis

Major:                CS

Faculty Adv:        Abby Marsh

Millions of users exist on the twitterverse, some human, some not. Can computers identify non-humans on twitter? We created an algorithm that grabbed users’ tweets and used them to classify users as either bot or human. In this algorithm, cosine similarity was used to gauge the similarity of a user’s tweets. The higher similarity, the more likely they came from a bot. In the end, the algorithm performed well, correctly identifying about 75% of users.


Effects of traumatic experiences on mental health in Ukraine

Student:        Charlotte Fowler

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

Ukraine has a long history of political conflict and turmoil, serving as the border between the West and the East. In this talk I will explore how Ukrainians’ exposures to traumatic experiences such as sexual assault or violence have increased their risk of developing Intermittent Explosive Disorder, an impulse control disorder.


Hearing Colors: Relationships between Albums and Color

Student:        Paige Pfeiffer

Major:                CS

Faculty Adv:        Getiria Onsongo

When listening to music, one might notice that a particular genre appears to contain album covers with a certain overall color or that certain colors are present in album covers during a particular season. By examining an album cover’s overall color, we can learn many details about the music’s genre, season, release date, tempo, and popularity. We will explore how album details from Spotify’s top songs can be visualized through image parsing and D3.


Efficient Algorithms for Approximating Matrix Functions

Student:        Tiffany Fan

Major:                MATH/AMS

Faculty Adv:        David Shuman

Imagine a heat map of the United States with erroneously reported and/or missing data. How can we fill in missing information or extract the general trend? One solution method relies on computations involving functions of large sparse matrices and vectors. Unfortunately, the exact computations can be extremely time consuming. I’ll present a new algorithm that uses ideas from linear algebra, numerical analysis, probability, and graph theory to efficiently approximate these computations.


Ray Tracing – Capturing the path of Light

Student:        Ojashvi Rautela

Major:                CS

Faculty Adv:        Ian Whitehead

In computer graphics, ray tracing is used to obtain high degree of visual realism in images and animations. It is a technique to visualize a virtual 3D scene by simulating real world light. Here, we implement a ray tracer to produce realistic lighting effects. Essentially, our algorithm traces the path of light from a virtual light source, simulating the way that the light scatters, reflects and refracts off the virtual objects defined on the screen.


A Computer Assistant to Write Cover Letters

Student:        Xinyu Yang

Major:                CS

Faculty Adv:        Susan Fox

Writing a cover letter can be the scariest moment when applying for a job. However, the process of writing doesn’t have to be scary when you understand the structure. Because of this, natural language processing and generation can help us generate cover letters. Come to the talk and see how the program picks up information from resumes and job descriptions, and put pieces into a complete cover letter.


Building voters’ preferences

Student:        Trung Nguyen

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

It's voting day and you are excited to practice your democracy rights! But wait, you have to vote on seven ballots at the same time. Even worse, you would like to see some of them work together rather than split up. For my project, I use a framework to formalize voting preferences and study properties of interesting structures, prompted by mathematical curiosity. Come see math meets politics!


Colorizing black-and-white photos in one click

Student:        Giang Duong

Major:                CS

Faculty Adv:        Leslie Myint

Have you ever wondered how old photographs would have looked like if color cameras had existed then? Realistic colorization of monotone images often takes weeks using Photoshop. Deep learning techniques, however, provide a “good guess” of the lifelike colors of dynamic objects such as people’s skin tones and clothes in just seconds. Come learn how AI decides colors for black-and-white photos, even Manga drawings, then explore the challenges and latest approaches to automatic colorization.


Saving the World Takes Time and Money

Student:        Gloria Odoemelam

Major:                MATH/AMS

Faculty Adv:        David Shuman

Imagine creating an invention that can change the world forever. To get credit for that and protect your idea, you’ll need to spend thousands of dollars and wait up to 3 years to be granted a patent. Does bureaucracy delay, or worse, bar innovation? In our study, we examine whether patents filed by individuals and small companies take longer to be approved by the U.S. Patent and Trademark Office (USPTO) than those filed by wealthier large corporations.


Signed, Scanned, Delivered: A Comparative Analysis of Sign Language Recognition Systems

Student:        Eric Frost

Major:                CS

Faculty Adv:        Shilad Sen

Translating words requires clear input, knowledge of different languages, and a speedy response. Unlike verbal communication, sign language uses physical features to express words while following these same processes. In my studies, I examine the efficiency of input, processing and runtime using different methods to determine the most effective translation combination moving forward.


Integrating Instruction in Game Design

Student:        Luke Brown

Major:                CS

Faculty Adv:        Paul Cantrell

Do you ever have to learn how to use new software but hate to read instructions? All software requires instruction on its use; however instruction can be integrated into the design to create intuitive “discovery” of software use. This talk will use a student-created game as a case study to demonstrate integrating instruction into design.


How Long Does A War Last?

Student:        Xinwei Huang

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

Why do some conflicts last for decades while others last for months? Do wars in Asia last longer than wars in Europe? What is the impact of other factors, like combat intensity, third party involvement, and the rate of escalation on the duration of a war? We use data about conflicts to answer these questions.


How much is a life worth?

Student:        Nikhita Jain

Major:                MATH/AMS

Faculty Adv:        Brianna Heggeseth

There have been lives lost from all countries due to kidnappings by terrorist groups. Yet a disproportionate amount of them have been Americans. According to Hostage U.S., out of 90 murdered hostages, 41 were Americans. However, Americans only represent 20% of hostages taken. This may be because the U.S. has a strict policy of not paying ransoms to terrorist groups. Does this policy increase time in captivity for Americans? Or are there other factors at play?


Predicting Music Genre Classifications: Music Signals and Machine Learning Models

Student:        Ariana Pooley

Major:                MATH/AMS

Faculty Adv:        David Shuman

From an introduction to signal processing to exploring different machine learning algorithms, this talk will discuss ways in which one can predict the music genre of an input signal.


Why Evictions Happen: A Survival Analysis on San Francisco Homes.

Student:        Joe Fish

Major:                MATH/AMS

Faculty Adv:        Lisa Lendway

The face of San Francisco has changed dramatically. As residents continue to be displaced, techies blame landlords, landlords blame city hall, and city hall blames the voters who block development. However, while there is no shortage of pointed fingers, there are fewer empirical answers of which factors have caused the displacement seen in San Francisco. Join this talk for answers about why San Francisco has changed and what other cities can do to not follow in its footsteps.


Facial Recognition (almost) in your pocket

Student:        Duc Le

Major:                CS

Faculty Adv:        Libby Shoop

Have you ever wondered how FaceID of iPhone or God’s Eye in the Fast and Furious series works? Generally, advanced facial recognition algorithms are complex to understand, and modern approaches using deep learning technique are too time-consuming and computing heavy to be replicated in the limited academic setting. My capstone project is an implementation of doing facial recognition in real-time by simplifying existing complicated algorithms, with a minimal cost in running time and computing power.


What makes a hit song: An analysis of songs' survival time on the Billboard Top 100 Chart

Student:        Jiaxuan Huang

Major:                MATH/AMS

Faculty Adv:        Kristin Heysse

Wondering what makes a hit song? Using a survival analysis framework, we investigate factors that affect how long a song stays on the Billboard Top 100 Chart. Come to see how the key, the loudness,  (even the danceability) can determine how often you will hear a song on the radio or how likely your dad will be unconsciously humming it.


I am a computer who understands users. AMA

Student:        Jonathan Scott

Major:                CS

Faculty Adv:        Shilad Sen

Search engines such as Google and virtual assistants such as Siri or Alexa have become centerpieces in our digital lives. They assist us in finding entertainment and in finding important information. But do these systems actually understand our intentions? This project applies supervised and unsupervised machine learning algorithms to tackle this problem.


Busted! Fake News Classifier

Student:        Tianyou Li

Major:                CS

Faculty Adv:        David Shuman

How can we identify fake news? The vocabulary used? The reliability of the source websites? Other features? My model takes into consideration both the words used in a news article and the network centralities of the source website within the whole Internet to predict whether an online news article is fake.


Can you spot it? The implementation of SMS spam messages detector using Machine Learning Techniques

Student:        Fouad El Hamdouni

Major:                CS

Faculty Adv:        Abby Marsh

Did you ever receive spam messages on your smartphone? Did you ever wish there was a better way to detect them? We implement a deep machine learning technique called Long-Short Term Memory (LSTM) and compare its performance to traditional Naive Bayes in filtering SMS spam messages. The LSTM model shows improvement in both performance and accuracy compared to Naive Bayes, providing a misclassification rate between 1% and 2%.


Global Peacekeeper - A Survival Analysis on War Duration

Student:        Xue Rui

Major:                MATH/AMS

Faculty Adv:        Lisa Lendway

Are you interested in international conflicts and global peacekeeping? Have you ever wondered why some wars just keep going on and on for decades? Does international involvement actually help to end conflicts quickly? In this talk, I will list out a few critical factors that significantly contribute to longer wars and establish a mathematical model predicting war duration using survival analysis. Join me if you are interested in an interdisciplinary experience of statistics and poli-science.


 Save or Change?: How different features affect the lifespan of a smartphone

Student:        Erle Lei

Major:                MATH/AMS

Faculty Adv:        Alicia Johnson

How often do you throw away your phone? Do you know that among the 9.4 million tons of electronic waste Americans produce each year, only 12.5% is recycled and the rest is simply dumped on those in developing countries? We conducted two behavioural surveys to study the factors that affect the lifespan of your smartphone. Join us to learn about the findings and how this might inform more sustainable behaviors.


Order in Chaos: Using Computational Topology to Study Biological Aggregations

Student:        Maitrayee Deka

Major:                CS

Faculty Adv:        Vittorio Addona

Biological aggregations such as schools of fish and swarms of insects are often governed by social interactions and complex, coordinated movement. What tools, then, can be used to study the evolving substructures within such formations? In this talk, I will explore how computational topology, coupled with machine learning algorithms, can be used to answer such questions.


How To Cover Your Tracks Using Signal Processing

Student:        Tianyou Li

Major:                MATH/AMS

Faculty Adv:        David Shuman

You stole a sandwich from a convenience store and got caught on the surveillance camera. What do you do? You can completely remove yourself from the video by analyzing and filtering the video using time-vertex signal processing. I will explain the method and demonstrate more applications including smoothing, denoising and inpainting graph signals and videos.


Virtual Chimes: User Interface Design for VR Environments

Student:        Brighten Jelke

Major:                CS

Faculty Adv:        Getiria Onsongo

Would you rather play a virtual instrument, or a physical one? One of the great advantages of virtual reality applications is that you can do things you'd never be able to do in regular old reality. In this talk, I'll walk through the process for building a playable musical instrument in VR.  First, I'll give some background on user interfaces and design principles, and then I'll show how those were implemented to create Virtual Chimes.


Popularity versus Similarity: Simulating Network Growth

Student:        Alice Qingyu Zhu

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

Popular YouTubers tend to get more popular, and there are simple network growth models that capture this “rich get richer” phenomenon. But if I don’t care about unboxing, then I’m not going to follow an unboxing YouTuber, no matter how popular they are. Taking this interest-sharing idea into consideration, we look at a network growth model that tracks both popularity and similarity, and explore its structural differences from a pure popularity model.


Which Flight to Pick? Factors of Flight Cancellations and Delays

Student:        Xuchen Wang

Major:                MATH/AMS

Faculty Adv:        Alicia Johnson

Plan your flight based on statistical science! We all have experienced flight cancellations and delays. They are annoying not only because of the endless waiting in the terminal but also because we have to rearrange all of our schedules accordingly. As statisticians, we are curious about what data can tell us about that. Come to this talk to learn which flights are preferable and which should be avoided!


"Shall I Compare Thee To Thy Nearest Neighbors?": Combining Semantic and Phonetic Relevance for Poem Generation

Student:        Anja Beth Swoap

Major:                CS

Faculty Adv:        Abby Marsh

If creativity is what makes us human, what does it mean if machines are able to paint pictures or compose sonnets? In this project, I explored methods to automatically generate pleasant-sounding poems with coherent topics. By combining similar lines, measuring similarity in terms of both meaning and sound, free-form poems can be written. The resulting poems are promising, and often intriguing, even if they do not yet approach human-level writing.


Single, Parallel, Graphical: Solving All Pairs Shortest Path through Parallel Processing

Student:        Eric Huang

Major:                CS

Faculty Adv:        Lian Duan

Are programs that are written for parallel processing always faster than their single-core processing? What kind of problems are difficult to parallelize? In this talk, I tackle a parallel runtime analysis on the All Pairs Shortest Path algorithm, which belongs to a class of problems historically difficult to parallelize, known as dynamic programming. Come see how single-core, multicore, and GPU processors fare against this problem, and who actually comes out on top!


The Gravity of Game Design

Student:        Emerald Thole

Major:                CS

Faculty Adv:        Getiria Onsongo

Gravitas is a 2D platformer where you must manipulate the gravitational pull of special objects in order to overcome obstacles. I worked with a small team to polish and release what was previously a final project for the Software Development class. In this talk I’ll show off parts of the game and talk about the work that went into getting the game released to the public.


Minecraft AI

Student:        Emmanuel Diaz        

Major:                CS

Faculty Adv:        Libby Shoop

This project is simply about making an AI to play the game Minecraft. Minecraft is a survival crafting game which encourages exploration and creativity which makes it a great playground for AI. If you want to see what my AI can do you should come to my talk! You'll also be able to learn some stuff about the field of AI and the steps I took in the creation of my Minecraft AI.


Is MIA Biased? How Our Favorite Art Museum Plays Favorites

Student:        Zachary Skluzacek

Major:                MATH/AMS

Faculty Adv:        Kristin Heysse

Art museums and galleries are often regarded as bastions of acceptance and forward-thinking. The public confidence in these institutions’ social awareness is infrequently challenged, and so we chose to statistically analyze the Minneapolis Institute of Art (MIA) to determine if it fulfills these expectations.


Maybe She's Born With It, Maybe It's The Shading Algorithm  

Student:        Sarah Pujol

Major:                CS

Faculty Adv:        Lauren Milne

From Peach’s blocky hair to Rapunzel's flowing locks, how do these animated characters’ hair come to life? After choosing the perfect hair model, a shading algorithm must be implemented to create depth and highlighting in the hair. The key to perfectly highlighted locks is an algorithm that mimics the physics of light interaction with hair. I will walk through the process of implementing the Marschner Reflectance Model that brought my hair model to life.


Survival Analysis in Capital Punishment: Death Row Inmates’ Sojourn time in Texas Prisons  

Student:        Kristijan Peev

Major:                MATH/AMS

Faculty Adv:        Tom Halverson

Texas leads the nation in the number of executions since the death penalty was reinstated in 1976. Once one is convicted of the death penalty, the time it takes for them to be executed is defined as sojourn time. Are there characteristics such as age, race, gender, educational attainment or past criminal record that affect how long a person stays on death row? Would the victim’s race and gender matter at all?


The Craft of Software Design: How to Give Your Code a Makeover  

Student:        Kelli Mandel

Major:                CS

Faculty Adv:        Paul Cantrell

Just because code works doesn’t mean it’s elegant. Disorganized code plagues software developers around the globe; adding features or finding bugs can be difficult when the code is unclear or jumbled. This is where refactoring comes in. But what is refactoring? How is it done and how does it help developers? This talk will introduce a better understanding of the importance of code organization and how to improve it in JavaScript.


Music Materialized: Graphically Enhancing the DJ Experience  

Student:        Ansel Colby

Major:                CS

Faculty Adv:        Libby Shoop

Many DJs' rely on prebuilt graphics that respond to their music in a fixed manner. In this talk I will explore how providing the ability to tweak and fine tune a graphically responding scene can be more effective than preloaded scenes, and discuss the process of how this project enhanced the role of the graphical DJ.


Are you changing your cell phone or changing the world?  

Student:        Zhen Tan

Major:                MATH/AMS

Faculty Adv:        Alicia Johnson

Why do people change their phones? How often do they do that? How can a proper disposal of a smartphone benefit not only big tech companies but also this planet? These questions inspired our investigation into the lifespan of smartphones. This investigation, based on the results from survival analysis, could help big tech companies develop effective marketing strategies or resolve the increasingly serious problem of electronic waste.


Deal or No Deal: Examining the Effects of Ransom Payment on Hostage Durations  

Student:        Maitrayee Deka

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

The United States has a well known ‘No-Concessions’ policy when it comes to dealing with terrorists. Although designed to deter kidnappings of American nationals, a disproportionate number of murdered hostages are from the US. Does this strategy of not negotiating ransom with violent groups actually help US citizens or hurt them? In this talk, I will explore how factors such as ransom payment and nationality affect the time to release of victims in hostage situations.


Playing with Fire: A Survival Analysis of Wildfire Size and Duration

Student:        Matt Yang

Major:                MATH/AMS

Faculty Adv:        Tom Halverson

Over 100,000 wildfires ignite in the US each year, thereby burning millions of acres of land and destroying many communities. But, only a handful of these incidents grow to become the devastating wildfires we see on the news. With worsening fire conditions, important questions arise regarding the best ways to prepare for wildfires and the best ways to respond. This project uses survival models to better understand the factors that lead to fire growth.


Building Better Poker AI Through Opponent Modelling  

Student:        Andrew Cui

Major:                CS

Faculty Adv:        Susan Fox

Hollywood portrays poker as a game of reading people: every detail counts. However, in a game of poker a gesture or a certain play could be coincidental. How confidently can we believe that the information we gained reflect our opponents’ strategy in a meaningful way? This problem is key to building more efficient poker AIs. In our project, we simulated poker game scenarios and explored different ways to infer opponents’ strategies from available information.  


God, I Hope I Get It! I Hope I Get It!: The Impact of the Tony Awards on Broadway Show Longevity

Student:        Kavya Shetty

Major:                MATH/AMS

Faculty Adv:        Brianna Heggeseth

Why is Phantom of the Opera still running? Why has no one heard of American Buffalo? Ever wondered why some shows have a seemingly permanent seat under the shining lights of Times Square while others die? The Tony Award is considered the most prestigious commendation a Broadway production can receive--but does the fancy golden statue guarantee success?


Digitizing, Districting, and Data: Evaluating Political Gerrymandering in Ohio

Student:        Katie Jolly

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

Many people claim that political districts are gerrymandered, but how do we provide quantitative evidence? Modern gerrymandering research has furthered computational methods to evaluate the fairness of political districting. My work aims to make that process more accessible for researchers and community members. In this talk, I will illustrate those methods by testing to see if current Ohio districts are gerrymandered as well as discuss some broader themes in civic engagement work.


First Moment of Quadratic L-functions in Function Fields

Student:        Dona Pantova

Major:                MATH/AMS

Faculty Adv:        Ian Whitehead

I will study a family of L-functions that carries fascinating number-theoretic information. The functions in this family are polynomials whose coefficients count square and non-square polynomials, and exhibit some interesting patterns related to the polynomial degree. Studying how fast L-functions grow is an important problem in analytic number theory. My project explores the growth of a family of L-functions by studying the values of the L-function averaged over the family.


Building 3D Virtual Chimes: Interactive Feedback in Virtual Reality

Student:        Kayla Beckham

Major:                CS

Faculty Adv:        Getiria Onsongo

Virtual Reality (VR) interfaces are most effective  when users can interact with virtual objects as they would in real life.  Our senses expect a response when we move or touch something. For VR to fool our senses, the feedback must match up to user actions. This talk will explore the process of implementing visual, auditory, and haptic (vibration) feedback in the development process for a 3D virtual musical instrument.


Fighting Twitter Clutter

Student:        Zachary Skluzacek

Major:                CS

Faculty Adv:        Kristin Heysse

Football? The Oscars? Video games? Whatever topic you want to block from your Twitter feed is probably being discussed on Reddit somewhere. Using machine learning, we can teach computers to recognize any of these trends to keep your Twitter feed free of clutter.


Timer of the Equalizer : Survival Analysis of the leading team in LaLiga

Student:        Haroon Malik

Major:                MATH/AMS

Faculty Adv:        Joe Benson

Have you ever wondered what factors matter the most when equalizing in a soccer game? Is it team quality, home field advantage or the match statistics? Join me in my presentation as I show the results of a survival analysis of Spain’s top flight competition to answer these questions.


Neural Network Techniques on Spam Text Detection

Student:        Xuchen Wang

Major:                CS

Faculty Adv:        Alicia Johnson

Have you ever received spam text messages? Have you ever wished your phone could get rid of those spam messages automatically? This talk combines newly developed neural network techniques with the traditional text classification problem. Come to this talk to learn about the basic background and the power of neural networks!  


Supercomputing for the Broke College Student

Student:        Amanda Doan

Major:                CS

Faculty Adv:        Libby Shoop

Have you ever wanted to analyze performance of an algorithm on a supercomputer, only to realize that buying one would cost $4 million+? Don't you fret! In this presentation, I'll talk about a project I did when I interned at Cray Inc., a supercomputing company, and methods of scaling it down to run on an affordable 10-node PicoCluster while maintaining performance.


Fire and Fury: A Survival Analysis of Wildfire Cause and Duration

Student:        Usman Hasan

Major:                MATH/AMS

Faculty Adv:        Tom Halverson

Wildfires pose an enormous threat to many parts of the United States and the world at large. With worsening climate change, the danger is only increasing. But what causes wildfires, and why are some fires so much worse than others? What can we do about it? To better understand this phenomenon, we employed survival analysis techniques to study their causes and behavior. I will discuss our findings and possible avenues for future research.


Citation "Queer"ies

Student:        Maggie Hohenstein        

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

The big names of Queer theory include Foucault and Butler, but are their works cited as often as they are talked about? How can we see some of the power structures that queer theory is critiqued for at play in a citation network of queer theorists? By connecting authors based on who they are cited by, we can easily see a visual representation of the community of queer theory and its critics.


You're Fired: On Soccer Head Coach Tenure Lengths

Student:        Troy Lyu

Major:                MATH/AMS

Faculty Adv:        Vittorio Addona

Between 1999 and 2015, an average soccer head coach leaves the team after only 2 1/2 years; a quarter of them couldn't last a year. What contributes to a head coach being fired earlier or later? Is it just the standings? Could nationality be a factor? How do different leagues compare?


Waiting in the Terminal: An Analysis of Flight Delays

Student:        Ruiqi (Annie) Wang

Major:                MATH/AMS

Faculty Adv:        Alicia Johnson

More than 1 million flights were delayed last year, which translates to nearly 64 million minutes worth of delays. Want to say goodbye to delayed flights for the rest of your life? Come to learn which airlines and days to fly and which to avoid. This talk uses survival analysis and machine learning techniques applied to 5 million domestic flights in 2015, to explore the factors that influence flight delay durations. You’ll be surprised.


Is Sociology missing something? Limitations of traditional network data collection

Student:        Ruotong Wang

Major:                MATH/AMS

Faculty Adv:        Ian Whitehead

The quality of data directly affects the quality of research. While social media such as Twitter or Facebook allow researchers to capture large scale complete social networks, smaller scale network data collection methods are still widely used in Sociology due to easy access. My capstone compares the characteristics of networks captured by online social media and by smaller-scale data collection methods and identifies the potential limitations of the traditional data collection methods.


Pac-Sphere: Pac-Man on a sphere

Student:        Gabriel Brown

Major:                CS

Faculty Adv:        Tom Halverson

Pac-Man: the myth, the puck, the legend. Thousands of players have sunk countless hours into this iconic game. But what if it were three-dimensional, and set on a spherical maze? That’s what I set out to make with Pac-Sphere. Come by if you want to learn about skeletal animation, displacement mapping, and the fascinating and often frustrating world of computer graphics.


Theatrobo: Exploring Robot Theater with Shakespeare

Student:        Andy Han

Major:                CS

Faculty Adv:        Susan Fox

Can you imagine a play at the Guthrie Theater performed by robots? My teammates and I programmed Theatrobo, a program that allows a LEGO robot to perform a scene from Hamlet alongside humans. I will provide an overview of both robot theater and Theatrobo. Lastly, I will conclude with my thoughts on the intersection of robotics, art, and society. Join me to discover a sub-field of robotics that’s not just self-driving cars or smart speakers.


Lose Yourself: Slim Shady does Combinatorics

Student:        Blake Jones

Major:                MATH/AMS

Faculty Adv:        Andrew Beveridge

If you’ve ever wondered just how lazy a combinatorialist can be, you’ve come to the right place to find out. We’ll talk about all essential aspects of the mathematician’s college experience, ranging from drunk walks home late at night, and figuring out how to get out of long winded calculations. Of course, along the way, we’ll uncover the shocking similarity between lattice paths and search trees, and how to completely lose yourself (rapping not included).


Modeling Realistic Human Movement

Student:        Mark Coretsopoulos

Major:                CS

Faculty Adv:        Libby Shoop

Have you ever wanted to know what separates rubbery, unconvincing animation from animation that mimics the muscles and bones of the human body? This talk will cover an animation technique called skinning, which dictates how a 3D model twists and flexes in response to movement. Our project takes an existing static model loader, and adds skinning and animation.


No need to Trust anyone: secure online voting system

Student:        Chen Dun

Major:                MATH/AMS

Faculty Adv:        Ian Whitehead

There is always a risk in trust, especially when we must trust the system in which it is highly profitable to cheat as in voting system. But when we vote, we need to trust the register of our identity, the counter to unanimously and accurately count our vote, and even other voters are following the rule of voting. In this project I will make a fully decentralized secure online voting system, trusting no one.


Changing Coordinates in Permutation Groups

Student:        Sam Armon

Major:                MATH/AMS

Faculty Adv:        Tom Halverson

This project presents a solution to an 80-year-old problem in representation theory, a field in which we represent a permutation of the numbers {1, 2, … , n} with a matrix. There are two well-known coordinate systems in which these representations can be presented, each revealing different information about the underlying permutations. We prove a general formula for the change-of-basis matrix between these two coordinates systems using combinatorics and walks on graphs.  


Avoiding the B-word: a Survival Analysis of Corporate Defaults

Student:        Carl Salvino

Major:                MATH/AMS

Faculty Adv:        Lisa Lendway

What do Tesla, General Electric and PG&E all have in common? Besides maintaining market capitalizations far north of nine figures, they dominated the 2018 editions of the Wall Street Journal with their persistent flirtation with defaulting and entering into the dreaded land of bankruptcy. Through this talk I will use methods of survival analysis to identify risk factors that lead to corporate defaults and compare their predictive ability to existing marketing indicators.


Simulated Jenga: Using Unity to Create a Video Game

Student:        Courtney Overland

Major:                CS

Faculty Adv:        Susan Fox

Have you ever wanted to create a video game but you didn’t know where to start? In this talk I will explain my journey in creating a simulation of the classic wood block game, Jenga. I’ll talk about the Unity software, consider various user interface techniques, discuss how to use and manipulate models and prefabs, and explain some of the challenges associated with creating a 3D game which is played in a 2D space.


Syllable Separation and Consonant Classification with Energy Transforms

Student:        Amanda Doan

Major:                MATH/AMS

Faculty Adv:        Libby Shoop

There are many different approaches to speech segmentation and the problem of SUV classification. In order to understand and evaluate blind speech segmentation and SUV (silent, voiced, unvoiced) classification, we implement two algorithms using a short term energy function and a frame- based Teager energy function, and evaluate their accuracy. This presentation is an overview of the concepts and these implementations.


???

Student:        Haolin Chen

Major:                ???

Faculty Adv:        Vittorio Addona

???


Bot Eye vs Human Eye: Understanding the Brain’s Visual System through Modeling V1 Striate Cells

Student:        Elena Smith

Major:                CS

Faculty Adv:        Abby Marsh

How do eyes determine the orientation of movement? To understand the visual cortex in a new way, we compare and contrast what is understood about the visual cortex to how AI techniques accomplish movement recognition. Using Computer Vision algorithms Harris Corner Detection and SSD tracking, we model the striate cells’ activation. This work informs the existing model of the visual cortex, providing a framework for how the visual information is extracted from the environment.


Bulls, Bears, and Black Swans, oh My!

Student:        Tim Lipman

Major:                MATH/AMS

Faculty Adv:        Lisa Lendway

Many think the only way to making money on wall street is finding the unicorn company that will change the way we live. My presentation is on something much better than unicorns: defaulting companies. While there is no glitter, predicting bankruptcy can provide an excellent opportunity and even match investments on unicorn companies. This presentation will use survival analysis for predicting company bankruptcy, and we will discuss distressed debt. I implore you to come to my talk.


Assignment Planning for the Visual Learner

Student:        Jeremy Andrew Krovitz

Major:                CS

Faculty Adv:        Libby Shoop

Have you ever wanted to visualize your assignments over a short term and a long term time period? Did you ever wish you could see your assignments laid out from start to finish and which ones you needed to get done first? Come see a talk about a web application that will revolutionize your planning needs and allow you to get all of your assignments completed on time!


                                                                                                                        Top