Monday, March 18, 2024
10:00am-12:00pm
RS208, Rosebrugh Building,
164 College St.
Quan Nguyen, University of British Columbia
Mock Lecture (50 Mins): A Technical Introduction to Large Language Models
Technical Talk (35 Mins) & Q&A (15 Mins): Cross-Institutional Transfer Learning in Student Dropout Prediction Models
Modern machine learning increasingly supports paradigms that are multi-institutional (using data from multiple institutions during training) or cross-institutional (using models from multiple institutions for inference), but the empirical effects of these paradigms are not well understood. In this talk, I will investigate cross-institutional learning by assessing the utility and fairness of student dropout prediction models that are transferred across institutions. I examine the feasibility of cross-institutional transfer under real-world data and model-sharing constraints, quantifying model biases for intersectional student identities, characterizing potential disparate impact due to these biases, and investigating the impact of various transfer learning approaches (i.e., direct, voting, stacked transfer) on fairness and overall model performance. The analysis was conducted on data representing over 200,000 enrolled students annually from four US universities without sharing training data between institutions.
Bio:
Dr. Quan Nguyen is working as a Postdoctoral Fellow in the Department of Statistics, University of British Columbia. He brings over 5 years of teaching experience in data science and machine learning in both undergraduate and graduate levels including the UBC Master of Data Science, UBC Master of Business Analytics, and the University of Michigan’s Master of Applied Data Science. Quan also oversees the UBC MDS capstone program by coordinating 20-30 data science projects in collaboration with industry partners across Canada. His research interests lie in the intersection of data science and education, with a focus on developing predictive models of academic success and student engagement with learning management systems. His contributions to the field of learning analytics have been recognized with 2 best paper awards and a grant from the Michigan Institute of Data Science.
Evaluation Questions Link: https://forms.office.com/r/PAGvZJ7YG5
