Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.
Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.
Author ORCID Identifier
https://orcid.org/0009-0005-8116-019X
AccessType
Open Access Dissertation
Document Type
dissertation
Degree Name
Doctor of Philosophy (PhD)
Degree Program
Computer Science
Year Degree Awarded
2024
Month Degree Awarded
February
First Advisor
Justin Domke
Subject Categories
Applied Statistics | Artificial Intelligence and Robotics | Other Applied Mathematics | Other Mathematics | Probability | Statistical Methodology
Abstract
Variational Inference performance strongly depends on many algorithmic choices, such as the optimization algorithm used, the objective, and the variational family. While the right combination for these components is highly problem dependent, there is currently little guidance on how to find it. This limits the use of these methods by non-expert users. This thesis advances Variational Inference towards an automatic and robust technology. We provide theoretically justified guiding principles for these decisions, and algorithms able to make optimal choices adaptively, creating Variational Inference based tools that are robust, flexible, and widely accessible.
DOI
https://doi.org/10.7275/36489445
Recommended Citation
Geffner, Tomas, "Towards Automatic and Robust Variational Inference" (2024). Doctoral Dissertations. 3050.
https://doi.org/10.7275/36489445
https://scholarworks.umass.edu/dissertations_2/3050
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
Included in
Applied Statistics Commons, Artificial Intelligence and Robotics Commons, Other Applied Mathematics Commons, Other Mathematics Commons, Probability Commons, Statistical Methodology Commons