Search

Blair Bilodeau
  • Home
  • CV
  • Papers
  • Teaching
  • Projects
Avatar

Blair Bilodeau

PhD Candidate in Statistics

University of Toronto

Vector Institute

About

I’m a PhD candidate in statistics advised by Daniel Roy at the University of Toronto, supported by the Vector Institute and an NSERC Doctoral Canada Graduate Scholarship. I received my BSc in financial mathematics from Western University in 2018.

My research area is broadly statistical machine learning, with a focus on theoretical performance guarantees for sequential decision making. Some problems I’m currently thinking about are:

Adaptivity: How can we characterize the difficulty of learning data beyond classical stationary dependence structures and design algorithms that adapt to these difficulty notions?

Statistical Complexities: Existing notions of predictor and algorithm complexity may be vacuous or suboptimal for modern machine learning systems. What are the right notions of complexity that lead to matching theoretical guarantees on empirical performance?

Transfer Learning: Empirical performance has far outpaced the theory of learning data with few or no labels and using this data to make predictions on out-of-distribution data. What are the correct theoretical formalizations of these tasks that result in sample complexity guarantees representative of empirical performance?

To view my curriculum vitae, click here.

Preprints and Articles

*denotes equal contribution

Relaxing the I.I.D. Assumption: Adaptively Minimax Optimal Regret via Root-Entropic Regularization

Blair Bilodeau*, Jeffrey Negrea*, Daniel M. Roy
2020 arXiv Preprint
Cite arXiv

Tight Bounds on Minimax Regret Under Logarithmic Loss via Self-Concordance

Blair Bilodeau, Dylan J. Foster, Daniel M. Roy
2020 International Conference on Machine Learning
Cite arXiv Poster Slides Camera Ready

Average Waiting Times in the Two-Class M/G/1 Delayed Accumulating Priority Queue

Blair Bilodeau, David A. Stanford
2020 arXiv Preprint
Cite arXiv Slides Code

Simulated Co-location of Patients Admitted to an Inpatient Internal Medicine Teaching Unit: Potential Impacts on Efficiency and Physician-Nurse Collaboration

Blair Bilodeau, David A. Stanford, Mark Goldszmidt, Andrew Appleton
2019 INFOR: Information Systems and Operational Research
Cite PDF Slides Code Camera Ready

Teaching

STA347: Probability Theory

Course Notes

Projects

Scripts to Scrape Data from arXiv and biorXiv

Slides Code

STA2201 Homework 3: Spatial Analysis Examples

Report

STA2101 Final Project: Spatial Analysis of Crime in Chicago

Report Code

© Blair Bilodeau 2021 · Powered by the Academic theme for Hugo.

Cite
Copy Download