Department of Mathematics, Duke University

Program ID: Duke-DATA2016 [#423]
Program Title: Data+ 2016
Program Type: Undergraduate program
Program Location: Durham, North Carolina 27708-0320, United States [map]
Subject Areas: Data Science, Interdisciplinary
Application Deadline: 2016/02/25 finished (posted 2015/12/16, listed until 2016/06/16)
Program Description:    

*** this program has been closed, and no new applications will be accepted. ***

Data+ is a ten-week summer research experience that welcomes Duke undergraduate and masters students interested in exploring new data-driven approaches to interdisciplinary challenges. Students join small project teams, working alongside other teams in a communal environment. They learn how to marshal, analyze, and visualize data, while gaining broad exposure to the modern world of data science.

Students work in small teams (at most 3 undergrads and 1 masters per team) on data-driven, interdisciplinary projects, but they do so in a communal environment with about 20 other teams working in an extremely diverse set of subject areas. It is our hope that students will be able to both work deeply into their specific project and get a very broad picture of most of the skills needed for modern data science.

Participants will receive a $5,000 stipend, out of which they must arrange their own housing and travel, as part of a Research Training Grant issued by the NSF to the Departments of Mathematics and Statistical Science at Duke, with additional funding and infrastructure support provided by the Information Initiative at Duke (iiD) the Social Science Research Institute (SSRI), Bass Connections, MEDx, and the Vice-Provost for Research. Participants may not accept employment or take classes during the program. The program runs from May 23 until July 29, 2016. The application deadline is Feb. 25, 2015, but we will evaluate applications on a rolling basis. We anticipate making our first round of offers in late January, so please get your applications in!

You will find the projects planned for summer 2016 in the numbered list below. Click on the project names to learn more. Please indicate the number of the project you choose when you apply. For some projects, IRB training may be required and will be provided in advance. With each project, we have attempted to list potential majors and/or interests that might be best suited for the project, but these should not be seen as requirements in any way! Quantitative STEM majors like mathematics, computer science, statistics, and electrical engineering are relevant to all.

1) Election Polling This project will explore new possibilities in political polling methodology, using data provided by a leading polling form. Sponsored by POLIS: The Center for Politics, Leadership, Innovation, and Service. Public policy, political science, all quantitative STEM.

2) Drugs and Gluttony How do individuals change their eating habits when given a new prescription? Working with the Sanford School and the BECR Center, students will analyze and visualize detailed food purchase data and prescription records to classify drugs, examine the effect of new prescriptions on food purchases, and measure the nutritional impact of any changes in food purchase behavior. Public health, economics, all quantitative STEM.

3) Data-Driven Parking Sponsored by the Parking and Transportation Office at Duke. Students will work with a dataset that carries de-identified information about parking behavior in the university and medical system. In consultation with professionals at that office, they will build vizualization and analytical tools to assist with strategic inventory management. Economics, all quantitative STEM.

4) Transgender Discrimination Survey Students will work with data from the National Transgender Discrimination Survey, collected as a joint project of the National Gay and Lesbian Task Force and the National Center for Transgender Equality. They will evaluate the consequences of anti-transgender bias on housing, employment, health, education, family life and criminal justice systems. Gender studies, public health, public policy, all quantitative STEM.

5) Smoking and Activity Space Students will use signal processing and network analysis to understand how the manner in which people allocate their lives across space and time is related to variables that affect their health. Using GPS activity tracks acquired from smokers and nonsmokers in Durham County, as well as demographic and purchase data about each subject. Economics, psychology, sociology, public health, all quantitative STEM.

6) Data-driven Development Sponsored by Alumni Affairs and Development at Duke. Students will investigate commonalities and distinctions in alumni gifts, and attempt to understand and predict motivations for gifts of different types. They will also construct mathematical models to evaluate different strategies for alumni engagement. Students will have the opportunity to consult with the Prospect Research, Management and Analytics team in the development office. Economics, Psychology, Sociology, all quantitative STEM.

7) Geometry and Topology for Data Sponsored by Geometric Data Analytics, Inc. Students will use methods inspired by geometry and topology on data relevant to vehicle tracking and/or cyber defense, all quantitative STEM.

8) Energy Resource Assessment How can satellite imagery be used to understand and analyze building energy efficiency, distributed solar generation, and the consumption of oil and coal? Using data from the U.S. Geological Survey, students will investigate methods and applications for using satellite imagery and aerial photography for energy resource assessment. Sponsored by the Energy Data Analytics Lab. Environmental sciences, economics, all quantitative STEM.

9) EMR and Clinical Trials This project involves the Triad Health Network and UNC Greensboro. Students will utilize electronic medical records data including medical codes, medications, demographics, lab values and vital signs, to develop a model for predicting future disease in patients with diabetes, build a scheme for randomizing high risk patients to either health coaching or control groups, and implement an approach to track cost and comorbidity outcomes in trial patients. Pre-med, public health, all quantitative STEM.

10) NC Budget Data and Policy This project is sponsored by the Budget and Tax Center, part of the North Carolina Justice Center. Students will help the BTC build a keystone tool for analysis of the North Carolina state budget, use historical budget data to run scenarios in order to illustrate the impact of proposed state-level policies, and help create a budget data visualization tool that allows the public to explore and learn about public investments in an interactive way. Public policy, political science, economics, all quantitative STEM.

11) Black Queen Hypothesis This project involves some travel to the Smithsonian National Museum of Natural History. Student will use image analysis techniques to test predictions about the relationship between body size and social complexity, in the context of the evolution of highly cooperative colonies of eusocial insects. Biology, philosophy, all quantitative STEM.

12) Fruit Fly Morphogenesis Students will analyze 4D images (3D + time) of cell movements and shape changes in developing fruit fly embryos to assess the effect of genetic mutations on morphogenesis. They will implement machine learning methods to detect and track cell boundaries in image sequences, and create a graphical user interface in which biologists may edit the results. Biology, all quantitative STEM.

13) Eye Movements and Food Choice Students will develop automated image processing algorithms to analyze mobile eye-tracking data, in an attempt to identify and code what subjects are looking at while they choose between different items in a mock “convenience-store” setting. Psychology, neuroscience, all quantitative STEM.

14) Durham Neighborhoods Sponsored by the Neighborhood Compass, a neighborhood indicators project designed to empower the communities of Durham County by leading and tracking services and action with data. Students will build visualization tools to explore and analyze datasets related to chronic ambient stress, population change, and energy consumption. There will be opportunities to present findings to city and county leadership. Public policy, economics, environmental science, history, public health, pre-med, sociology, all quantitative STEM.

15) Health Networks and Disparities Sponsored by the Duke Network Analysis Center. Students will build an interactive “co-treatment” network visualization tool, where each presented medical condition is linked to the other conditions most commonly associated with it in similar patients. This will help medical professionals see beyond the immediate condition that patients present and show how medical problems co-occur. Disparities by race, gender and poverty status will be analyzed. Sociology, public health, pre-med.

16) Night Vision Students will explore the genetics of night vision and darkness adaptation. The team will perform time series analysis of visual acuity and contrast sensitivity - night vision score measurements taken over a period of 20 minutes as participants adapt to the dark. Students will then evaluate the association of these temporal trends of darkness adaptation with genetic variant data. Biology, biostatistics, pre-med, all quantitative STEM.

17) Team Science Sponsored by a Clinical and Translational Science Award from NIH. Students will develop metrics for evaluating the amount of team-based science that occurs at Duke, and will produce a visualization map of team science in the Duke Schools of Medicine and Nursing. Pre-med, sociology, lab sciences, all quantitative STEM.

18) Predicting Pancreatic Cancer Students will electronic medical records (EMR) data to search for precursors of pancreatic cancer. They will assess whether patients with type II diabetes are at higher risk for pancreatic cancer, and develop a statistical model of demographic and medical record data. Biostatistics, pre-med, biology, all quantitative STEM.

19) LungMAP The LungMAP project seeks to improve lung health by providing the research community with a comprehensive web-based atlas to support investigations into the processes that regulate lung development. Students will develop a statistical and machine-learning pipeline to automatically classify immunofluorescent images of developing mouse lungs in the LungMAP database. Biostatistics, biology, pre-med, all quantitative STEM.

20) National Asset Scorecard for Communities of Color The NASCC is an ongoing survey project that gathers information about asset and debt positions of households at a detailed racial and national origin level. Students will work with Duke's Samuel DuBois Cook Center on Social Equity to use the survey data to examine various dimensions of social inequality including labor market discrimination, health outcomes, family structure, differential reliance on predatory lending sources, exposure to poverty, immigrant remittances, and equity in homes versus other types of assets. Public policy, economics, sociology, all quantitative STEM.

21) Smart(er) Routing at Theme Parks Students will have access to historical queue length and realized wait time data from several major theme parks, with the data provided by TouringPlans.com. The team will investigate the behavior of customers to determine the strategies and degree of sophistication that customers employ when deciding which rides to ride and when. Students will also seek to identify opportunities to shift customer behavior in a socially desirable way, with the goal of reducing congestion and improving the customer experience.

22) RTI-Duke Data+ Practicum Students will work with an enormous "synthetic population" representation of the United States, provided by sponsors RTI International. The team will develop a model that identifies locations for immunization and vaccination clinics in a large metropolitan area. Public health, Applied Mathematics, all quantitative STEM.

23) Faculty Research and Faculty Teaching This project will undertake a systematic investigation of the relationship between the research productivity of tenured and tenure-track faculty and the quantity and quality of teaching, using data from a highly prestigious research university.

24) Diagnosing Diabetes and Predicting Complications Sponsored by Northrup Grumman. Students will work with electronic health record data to develop predictive models of diabetes. Potential research questions include identifying patients with undiagnosed diabetes and predicting severe disease complications.

25 Learning to Search More DeeplyDifferent communities often pursue very different strategies for creating content on mobile platforms. Since the algorithms that train internet search tools are often not designed with a wide range of cultural competencies in mind, their output is often effected by problems of sampling bias, and thus is most useful only to a narrow set of end-users. A team of students from Duke and North Carolina Central will explore how transfer-learning techniques can be used to address these issues, and will construct prototype solutions. Students will work closely with the founder of this project's sponsor: Sankofa, Inc., a tech company making a difference for minorities in mobile.


Application Materials Required:
Submit the following items online at this website to complete your application:
And anything else requested in the program description.

Further Info:
http://www.math.duke.edu/mathbio/
 
Mathematics Department
Duke University, Box 90320
Durham, NC 27708-0320

© 2017 MathPrograms.Org, American Mathematical Society. All Rights Reserved.