Projects
Schedule Optimization for GO Transit Networks
Sept 2025 – Dec 2025
University of Waterloo (Waterloo, ON)
I formulated a mixed-integer programming (MIP) model to improve GO Transit transfer coordination by minimizing total passenger transfer waiting time, while enforcing capacity limits and minimum transfer-time constraints to keep solutions realistic. I processed GO Transit schedule data and built feasibility filters in Gurobi, including estimates for transfer demand, vehicle capacities, and allowable schedule adjustments so the optimization stayed implementable.
Result: the final optimized schedules showed that modest timing adjustments could reduce total peak-period passenger waiting time by 50%+, improving reliability for commuters without adding new infrastructure.
Deliverable: full optimization model + project report (methods, constraints, results, interpretation).
Skills/Tools: Python · Gurobi · MIP/Integer Programming · data processing · operations research · transit scheduling
Multiple Linear Regression Report for S&P 500 Price Dynamics
Sept 2024 – Dec 2024
University of Waterloo (Waterloo, ON)
I analyzed the relationship between the S&P 500 index level and financial / sector / macro variables using monthly data (1994–2023) sourced from the Bloomberg Terminal. I built and compared multiple regression specifications in R, addressing multicollinearity and model validity using correlation analysis, VIF diagnostics, and residual checks, and tested interaction effects where appropriate.
Result: I selected a best-performing model using Adjusted R², AIC, BIC, and PRESS, and clearly documented limitations driven by time dependence and non-stationarity (why strong in-sample fit can still fail in predictive settings).
Deliverable: written report with model selection workflow, diagnostics, and interpretation of findings/limitations.
Skills/Tools: R · Bloomberg Terminal · multiple linear regression · model diagnostics · AIC/BIC · VIF · financial data analysis