Student Projects

Photo of GW Data Science Capstone

The goal of the Capstone Project is to apply theoretical knowledge gained during the time at Data Science program into a realistic project that involves real datasets. During the project, students are heavily involved in the process of finding areal-world data science problems and solving it. The Capstone Project begins from collecting data and processing it in order to implement the appropriate analytic methods that they learned in the program to the real-world problems. In this process, problem statements and definitions play a major role in the Capstone and the datasets can be collected from industry, government, non-governmental organizations (NGOs), or academic research.

Learn more about some of the recent capstone projects completed by students from Data Science Program.

Capstone - Fixed Wireless Broadband

Capstone Completed By: Brian Wilson '20

Map of the US showing data from Brian Wilson's presentation

The FCC's Connect America Fund (CAF) is a reverse auction that aims to expand broadband connectivity to underserved regions of the US through grants to internet service providers. Connecting households via coax, fiber, and similar physical technology is prohibitively expensive. Wireless technologies make connectivity more economical but are limited to areas with minimal natural or man-made structures impeding signal propagation.

The goal for this project is to analyze all CAF-eligible census blocks to identify those most suitable for fixed wireless broadband deployment. Each block will be scored based upon housing density, tree coverage, and the area’s topology using publicly available geospatial datasets.

Access Brian's Capstone PowerPoint.

Capstone - Predicting Traffic Accident Risk and Severity 

Chart from presentation by Cristina Giraldo and Tashia Ferguson

Capstone Completed By: Cristina Giraldo '20 and Tashia Ferguson '20

Traffic accidents are responsible for 1.25 million deaths worldwide and is the fifth leading cause of death in the US (Global Status Report on Road Safety 2015, 2015). Injuries and costs associated with traffic accidents have a large impact on the individuals involved and the communities they live in. While there is a lot of current research on different modeling techniques in the Machine Learning and Deep Learning disciplines, there is still a need for further exploration in how these modeling techniques compare to one another when predicting accident risk and severity. This project aims to test several machine learning, time series, and deep learning models in order to decipher which modelling techniques have the best outcomes for predicting traffic accident risk and severity.

 Access Christina and Taisha's Capstone PowerPoint and Final Report

Capstone - Advancing the Course of Tele-Medicine with Artificial Intelligence 

Word cloud from Elie Tetteh-Wayoe's presentation

Capstone Completed By: Elie Tetteh-Wayoe '20

Tele-health or tele-medicine is an emerging area in health care delivery, yet there is no form of assessment to provide feedback to health care professionals on the quality of service rendered during their sessions. This affects the performance of these professionals and could have damaging effects on patient.

A system to assess the performance in a form of facial emotion recognition and speech evaluation to provide feedback in a timely fashion would help a great deal in the improvement of the quality of health care delivery. This research provides the techniques which are very crucial to the overall enhancement of tele-health.

Access Elie's Capstone PowerPoint and Final Report.