Declan McNamara
PhD Candidate in Statistics
University of Michigan
I’m a 5th year Ph.D. candidate in statistics at the University of Michigan, advised by Dr. Jeffrey Regier. My research focuses on methods and applications for scalable approximate Bayesian inference. I’m grateful to be supported as a Graduate Research Fellow of the National Science Foundation (NSF).
Interests
- Amortized inference
- Simulation-based inference
- Sequential Monte Carlo methods
Education
-
Ph.D., Statistics, 2025 (expected)
University of Michigan
-
B.S., Computational & Applied Mathematics, 2020
University of Chicago
-
B.A., Statistics, 2020
University of Chicago
Research
(2024).
Globally Convergent Variational Inference.
In Proceedings of The 38th Annual Conference on Neural Information Processing Systems (NeurIPS).
(2024).
Sequential Monte Carlo for Inclusive KL Minimization in Amortized Variational Inference.
In Proceedings of The 27th International Conference on Artificial Intelligence and Statistics (AISTATS).
(2024).
Variational Inference with Coverage Guarantees in Simulation-Based Inference.
In Proceedings of The 41st International Conference on Machine Learning (ICML).
(2023).
Simulation-Based Inference for Detecting Blending in Spectra.
In Workshop on Machine Learning and the Physical Sciences (NeurIPS 2023).
Teaching
2022 & 2023 M.S. Statistics Summer Bootcamp Instructor
Designed curriculum and taught three-day mathematics bootcamp to incoming Master’s students in statistics and data science programs covering topics such as multivariable calculus, linear algebra, and probability.
Undergraduate Research Program in Statistics (URPS) Mentor
Over the course of a semester, advised undergraduates in the completion of a research project applying deep learning and statistical methodology and to estimation predicting redshifts for objects in astronomical surveys such as SDSS.
Graduate Student Instructor (STATS 601)
PhD level course for first year students covering topics such as dimensionality reduction; factor analysis; classification; latent variable models; boosting; kernel-based methods; neural networks
Grader (Numerical Linear Algebra)
Computational course in linear algebra for statistics majors, covering topics such as eigenvectors, rank, nullity, SVD, QR, and other matrix decompositions.
Grader (Accelerated Analysis III)
Third quarter course in accelerated analysis sequence for mathematics majors, covering topics such as differential forms, Stokes’ theorem, and Lebesgue measure.
Teaching Assistant (Calculus I-III)
Led small group tutorials twice weekly for 6 undergraduate students for the non-mathematics major calculus sequence, covering topics such as limits, ε − δ proofs, univariate and multivariate differentiation and integration.