Master’s in Applied Data Science Autumn 2023 Capstone Winners

As the culmination of the Master of Science in Applied Data Science program, the Capstone Project unites students with industry partners to solve real-world analytics problems. At the Autumn Showcase—one of the largest in the history of the program—projects spanned industries, data types, and methodological approaches. Seven teams stood out for special commendation.

“The hard work, perseverance, and talent is clear from the truly impressive quality of the work,” said Greg Green, executive director for the program. “I was especially pleased to see how inventive the teams were in the solutions they developed for their industry partners.”

Learn more about the projects that won Best in Show below.

Brief AI | Revolution in Finance

Presenters: Jose Gerala Palacios, Vishal Parameswaran, Sanchit Narayan Kumar, Garima Sohi

Faculty Advisor: Nick Kadochnikov

Project Topic: Improved efficiency for financial analysts with AI tools to analyze earnings calls. The AI tools included a fine-tuning LLM trained on earnings calls to extract Key Performance Indicators and JAVI, a chatbot that allows you to have a conversation based on any earnings call.

Argonne National Labs | Smart Electric Vehicle Charging: A Reinforcement Learning Perspective

Presenters: Tim Fan, Suenkei Chan, Salman Yousaf, Keerthana Adavelli

Faculty Advisor: Gizem Aydin

Project Topic: Optimized EV smart charging to minimize grid strain by using a simulated environment and dataset based on Argonne’s EV charging infrastructure. “The goal was to train and tune a reinforcement learning framework and then evaluate it in terms of energy efficiency, peak power constraints, and user satisfaction,” the team said.

New Avenues | Increasing Efficacy of Traditional Employer-Sponsored Mental & Behavioral Support

Presenters: Jinghong (David) He, Luxin (Lucy) Huang, Xueyao Wang, Zhen Jia (Sunny) Sun

Faculty Advisor: Wendy Klusendorf

Project Topic: Developed a procedure for New Avenues that accurately matched employees and their dependents within their Employee Assistance Program to appropriate mental healthcare providers. The team did this by using data-driven machine learning to develop better practices for matching patients with providers by identifying the key patient symptoms that lead to a higher level of care.

Komatsu | Performance Evaluation of Komatsu Distributors with Telematics Data and Commodity Sales History

Presenters: Minh Vo, Maggie Chuang, Kelsey Liu, Sakshi Shende

Faculty Advisor: Roger Moore

Project Topic: Komatsu is a leading global manufacturer of construction, mining, and forestry equipment, and the capstone team used machine learning models  to analyze data and improve performance evaluation of Komatsu distributors. The team also leveraged commodity sales and the telematics machine data to identify relationships between machine data and parts consumption.

Research (Healthcare GPT) | AI Powered Radiology Report Automation

Presenters: Mariam Adeyemo, Saloni Khandelwal, Preetika Parashar, Yun Xing, Jeniffer Dongha Lee

Faculty Advisor: Utku Pamuksuz

Project Topic: Used AI to generate radiology reports based on findings and patient information using a large language model (LLM) trained on MRI and ultrasound data from UChicago Medicine. “This will greatly assist radiologists by advancing the automation of the documentation process,” the team said.

Research (Healthcare GPT) | Decoder-Only Computed Tomography Radiology Reports (DOCTRR)

Presenters: Tegan Keigher, Mary Erikson, Andrew Alvarez

Faculty Advisor: Utku Pamuksuz

Project Topic: Harnessed the potential of large language models to automatically generate accurate findings sections of CT radiology reports as this is a challenging task demanding substantial time and effort from radiologists. “Our vector-to-sequence model was in the end capable of producing radiology reports with a level of accuracy and quality comparable to that of human physicians,” the team said.

US Ignite | US Ignite National Broadband Tool

Presenters: Kishor Kumar Reddy, Shweta Sampath Kumar, Swathi Ganesan, Marc Edwards

Faculty Advisor: Don Patchell

Project Topic: Developed AI tool to predict economic impact of broadband access and write data- and insight-driven grant proposals to apply for and secure government funds. The capstone sponsor works with communities, businesses, local governments, and federal agencies to aid community leaders in securing funding and effectively managing broadband transformation projects. Existing studies have shown that nearly one-third of U.S. cities lack access to fast, reliable internet and that closing this gap would improve economic opportunities in those areas while also improving the quality of life for their populations.

 Honorable Mentions

ASL Recognition with Drone Technology

Students: Trang Dang, Josh Park, Kajal Shukla, Davis Thomas

Faculty Advisor: Steve Barry

Hedwig.ai

Students: Scott Matsubara, Kshitij Mittal, Radhika Sharma, Sunvid Aneja

Advisor: Nick Kadochnikov

 Utilization of Drone Tracking for Dog Search and Rescue

Students: Sanchin Arora, Indu Shekar, Anabella Liang, Priya Suvvaru Venkata

Faculty Advisor: Steve Barry

Leveraging Pre-Trained Large Language Models to Generate X-Ray Radiology Reports

Students: Josiah Chung, Ashlyn Fentem, Maria Fionalita, Isabella Xue, Murat Ozcan

Faculty Advisor: Utku Pamuksuz

KYC 2020 | KYC 2020 – Adverse Media Detection

Students: Sam Ding, Han Jiang, Richard Yang, Danhong Huang

Faculty Advisor: Nick Kadochnikov

 Mondeléz | Game-Theoretical Approach to Marketing Strategy in a Dynamic Competitive Market

Students: Devdutt Sharma, Suyash Lakhani, Parth Bansal, Irem Pamuksuz

Faculty Advisor: Anil Chaturvedi

Aerial Robotics for Liquid Level Detection and Inventory Management

Students: Zeyu Chen, Nicolás Ferreira, Peter Ryan, Chen Feng Tsai, Mia Zhang

Faculty Advisor: Steve Barry