Georgia Institute of Technology
Generalization in Dynamic Environments: Foundations and Algorithms
Abstract
dc:description.abstractDespite the success of modern machine learning models in a vast number of data-rich domains, an incomplete understanding of their robustness limits deployment in high-consequence scenarios. In particular, generalization often suffers as real-world conditions deviate from training environments and the strict assumptions of classical theory no longer apply. This thesis represents a joint theoretical and empirical investigation of generalization in dynamic environments. Its core tenet is that an understanding of the underlying mechanisms of generalization phenomena can elicit elegant methodological improvements. Therefore, we place equal emphasis on foundations — mathematical or experimental analyses which divulge fundamental insights into model behavior — and algorithms — adaptive methods which ensure model trustworthiness via rigorous benchmarking or theoretical guarantees. We study two different instantiations of dynamic environments. First, we investigate spurious correlations: patterns which are predictive of the target class in the training dataset but irrelevant to the true classification function. We present an empirical analysis of popular algorithms for the mitigation of spurious correlations across vision and language tasks, and we propose an adaptive method which greatly improves robustness despite requiring far fewer annotations. Then, we theoretically characterize feature learning dynamics of two-layer ReLU neural networks in a Boolean data setting where the spurious correlation is less complex than the quadratic target function. Second, we introduce a notion of task shift, wherein the objective function at test-time differs from the training objective. We provide a theoretical examination of this phenomenon in an overparameterized linear regression setting where the task shifts from classification during training to regression during evaluation.
Degree
thesis:*- Name thesis:degree_name
- Machine Learning, PhD
- Department thesis:degree_department
- Industrial and Systems Engineering
- Grantor
- Georgia Institute of Technology
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- LaBonte, Tyler
- Advisor dc:contributor.advisor
-
- Muthukumar, Vidya
- Committee members dc:contributor.committeemember
-
- Abernethy, Jacob
- Romberg, Justin
- Mussmann, Steve
- Goel, Surbhi
- ORCID person:identifier.orcid
Rights
- Language dc:language.iso
- English
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/81764