UW-Madison Research Bazaar 2020

- Michael Ferris, Director, Data Science Hub, Wisconsin Institute for Discovery
- Brian Yandell, Director, American Family Insurance Data Science Institute
- Tom Erickson, Director, School of Computing, Data, and Information Sciences

Discover, Collaborate, and Empower 

These three minute talks will allow the community to hear about 1) new and existing data science resources on campus, 2) data science applications for researchers in a variety of research domains, and 3) requests for new data science research collaborations. At the end of the session, participants will have the opportunity to talk with presenters to further discover opportunities, establish new collaborations and empower data rich research.

Lightning Talks in Round 1

Social Science Computing Cooperative: Resources for Data Science 
Russell Dimond, UW Social Science Computing Cooperative

Recover Analog and Digital Data 
Dorothea Salo, UW Information School

Machine Learning for Medical Imaging
Diego Hernando, UW Radiology and Medical Physics

What Works for Health: A Resource for Decision Makers 
Amanda Gatewood, UW Population Health Institute

Undergraduate Research with the Informatics Skunkworks
Dane Morgan, UW Informatics Skunkworks

Investigating Why Some Foreign Words Are Harder to Learn Than Others Using Duolingo
Elise Hopman, UW Psychology

Building a Community of Practice
Nathaniel Schleif, UW Plant Breeding and Genetics

Sharing Resources: Introduction to Semantic Technologies
Nancy Wiegand, UW Space Science Engineering Center

National Health Ambulatory Medical Care Survey
Ross Andrew, UW School of Medicine and Public Health Department of Radiology

Discover, Collaborate, and Empower 

These three minute talks will allow the community to hear about 1) new and existing data science resources on campus, 2) data science applications for researchers in a variety of research domains, and 3) requests for new data science research collaborations. At the end of the session, participants will have the opportunity to talk with presenters to further discover opportunities, establish new collaborations and empower data rich research.

Lightning Talks in Round 2

Using Machine Learning to Automate the Discovery of Microbial Viruses from Sequencing Data
Kristopher Kieft, UW Bacteriology

Using Citizen Data Successfully
Phil Henrickson, AE Business Solutions

GeoData@Wisconsin and the B1G Academic Alliance Geoportal
Jaime Martindale, UW Geography

Is the Public Cloud a Destination for Your Research? 
Joe Johnson, UW Division of Information Technology

Leveraging Software Development  and Data Science to Make Your Research Have More Impact
Steven Wangen, UW Wisconsin Institute for Discovery

Efficient Data Collection for Optimization
Yuji Saikai, UW Agriculture and Applied Economics

Deep Learning for Weather Satellite Imagery
John Cintineo, UW Space Science and Engineering Center

Creating a Culture of Cooperation 
Steve Goldstein, UW Botany

Resource: Center for High Throughput Computing
Lauren Michael, CHTC

Discover, Collaborate, and Empower 

These three minute talks will allow the community to hear about 1) new and existing data science resources on campus, 2) data science applications for researchers in a variety of research domains, and 3) requests for new data science research collaborations. At the end of the session, participants will have the opportunity to talk with presenters to further discover opportunities, establish new collaborations and empower data rich research.

Lightning Talks in Round 3

What Every Data Scientist Needs to Know about Cybersecurity
Bob Turner, UW Office of Cybersecurity

Data Services - Supporting Public Library Stories
Jody Hoesly, South Central Library System

County Health Rankings: Using Data (Science) for Action
Courtney Blomme, UW Population Health Institute

Technology for Social Change
Harshita Bhatt, Coding for Good

Engineering Proteins Faster Using Machine Learning Driven Method: MLProScape
Sanjan Gupta, UW Chemical and Biological Engineering

Applying Cloud Technology to Enhance Public Health Pathogen Surveillance
Kelsey Florek, WI State Lab for Hygiene

Geospatial Datascience Lab @ UW Madison
Song Gao, UW Geospatial Data Science Lab

Research Opportunities with the Survey of the Health of Wisconsin 
Jacquie Cronin, Survey of the Health of WI

The Environmental Data Initiative’s Approach to Environmental Data Curation and Publishing
Corinna Gries and Colin Smith, UW Environmental Data Initiative

Government agency staff have the knowledge and skills to analyze data using complex techniques like natural language processing and machine learning. We can use this data to predict which zip codes will have the highest mortality rates, or the most crime. However using these results can have far more drastic implications for the public sector than in the private sector.

I would like to bring together data scientists from state agencies to talk about how to ethically make predictions using our state data in a way that insures we are appropriately handling the data that our constituents allow us to use.

This discussion would support data scientists to appropriately use available data and techniques to help Wisconsin residents, providing caution and optimism for how state agencies can use new technologies for state improvements.

Openness and sharing of information are fundamental to the progress of science and to the effective functioning of the research enterprise. The free flow of information enables the research community to scrutinize scientific claims and to gain confidence over time in results and inferences that have stood up to repeated testing. Central to these topics are good data management practices that elevate data to first class research products that are citable and are attributed to creators. This workshop focuses on data management best practices occurring throughout the research life cycle and will focus on:

- Data management planning
- Collecting data
- Storing and sharing data during the research project
- Data quality management
- Metadata
- Data curation and publishing

Bring your questions so we can address all your data management challenges and use cases, and consider bringing a dataset. We find learning within the context of your own data is helpful.

This workshop will be an introduction to generating features for machine learning models from text. Text analytics, ranging from simple text features to more complex ones learned by neural networks, is a highly effective means of taking advantage of freeform written responses that researchers collect in surveys, scrape from the internet, or gather in the lab. These methods are therefore highly relevant to researchers at the Bazaar. Working with text data is also an invaluable skill in industry data science positions. The workshop will cover an introduction to different ways of characterizing text data, to extracting simple features from text data (e.g. bag-of-words), to more complex features that have proven themselves useful over time (e.g. word embeddings). We will cover a real-world application of these methods, focusing on a document classification model. Some knowledge of Python or another programming language will be useful. This workshop will use Python in Jupyter notebooks and will be platform-agnostic.

Note to workshop attendees:
Some exercises will be in Python 3, so attendees who wish to follow along with the provided exercises during the workshop should install nltk,scikit-learn, numpy, pandas, and scipy, ideally within a virtual environment (virtualenv or conda). Optionally, attendees may install thetransformers package. We will install additional packages during a designated break.

Qian (Jane) You, American Family Insurance 
Maureen Smith, Health Innovation Program 
Amy Yang, SuccessWorks
Tobin Magle, Division of Information Technology (DoIT)
Lauren Michael, Center for High Throughput Computing
Heather Shimon, UW Libraries
Martha Borgwardt, Epic