Open Data Science Conference(ODSC)

Boston, May4-5, 2017

Hynes Convention Center

Kapeel Chougule, WareLab

Conference snapshot

• Theme: applied data science conference covering latest trends in AI and data science topics.
• Over 4,500 participants, 20 training sessions, 27 workshops, 130 Speakers.
• a mix of several data themed talks covering application of AL, ML and DL, data science practices , and latest data visualization tools, among others
• enlightening sessions, workshops and booths
• I had conversations with folks from DataRobot, AWS, RStudio, R-brain, Microsoft and plenty more!

​

Packed house!!! ~4500 participants

Favorite Keynote talk

Four Eras of Analytics and Data Science from Artisanal Analytics to Augmented Automation

-

Thomas Davenport

Babson College/MIT fellow

​

​

​

Favorite session

• Programming language for data science.
• Started in MIT, 2012
• > 13000 registered packages
• Designed for parallelism and distributed computing
• As fast as C and as nice Python/R
• Rich descriptive datatypes and multiple dispatch.
• The demo showed the datatypes in Julia and parallelism functions
• JuliaDB workflow
• iJulia- Jupyter interactive environment with Julia

Artificial Intelligence, Machine Learning and Deep Learning

Artificial Intelligence(AL):

AI can refer to anything from a computer program playing a game of chess, to a voice-recognition system like Alexa, any m/c

​

Machine Learning(ML):

is one subfield of AI. The core principle here is that machines take data and "learn" for themselves.

​

Deep Learning(DL):

Deep learning is a subset of ML. It uses some ML techniques to solve real-world problems by tapping into neural networks that simulate human decision-making

Scalable Deep Learning with Apache MXnet on AWS- Dan Mbanga- Sr. Technical Manager- AWS

Apache Spark

• is a fast, in-memory data processing engine with elegant
• It provides in memory computations for increased speed and data process over mapreduce
• Sits on top of Hadoop cluster
• Spark is a library that enables parallel computation via function calls
• Features:
• Speed
• Ease of use
• Combine sql, streaming
• Run on diverse platforms
• Applications
• Iterative Algorithms in Machine Learning
• Sensor data processing
• Stream processing

Keynote talk: Matei Zaharia, Chief Technologist at DataBricks

Favorite workshop

Machine learning for Beginners- Hadelin de-Ponteves- creator of machine learning course on Udemy

Workshop covered

• Into to three algorithms- Naive Bayes, Support Vector Machine(SVM) and Thompson Sampling
• Application using some simple real world examples

R and Visualization session

• R and Bioconductor in Cancer Research- Big data in Genomics
• Aedin Culhane, Dana Farber Cancer Center,
• Showed how to use R and Bioconductor packages and cloud infrastructure to integrate analyze multiple data of genomics over 10,000 cancer patients
• Data Science lifecycle with Apache Zepplin
• Moon SooLee, NFLABS
• Showed interactive data analysis and visualization use Zepplin notebooks
• Create and preprocessing a design matrix with recipes-
• Max Khun, Software Engineer at RStudio
• New package called recepie for preprocessing data.
• Intro to D3 visualization
• Interactive Plotting with MatPlotlib
• Rapid prototyping data products using R-Shiny

Favorite booth

R-brain

RStudio

Microsoft

Favorite Quote

Conclusion

• I will give it a 4.5/5.
• Downside, multiple session going concurrently, so missed some important talks
• Great talks, good coverage on the latest trends in data science and AI
• It helped to access the current set of technology I am using to data science and learn the new trends and implement them with projects I am working on.
• Definitely recommend for those interested in Data science

​