Year: 2024
Type: preprint
Abstract: We study the problem of domain adaptation under distribution shift, where the shift is due to a change in the distribution of an unobserved, latent variable that confounds both the covariates and the labels. In this setting, neither the covariate shift nor the label shift assumptions apply. Our approach to adaptation employs proximal causal learning, a technique for estimating causal effects in settings where proxies of unobserved confounders are available. We demonstrate that proxy variables allow for adaptation to distribution shift without explicitly recovering or modeling latent variables. We consider two settings, (i) Concept Bottleneck: an additional ''concept'' variable is observed that mediates the relationship between the covariates and labels; (ii) Multi-domain: training data from multiple source domains is available, where each source domain exhibits a different distribution over the latent confounder. We develop a two-stage kernel estimation approach to adapt to complex distribution shifts in both settings. In our experiments, we show that our approach outperforms other methods, notably those which explicitly recover the latent confounder.
Source: arXiv (Cornell University)
Authors: Tsai, Katherine, Pfohl, Stephen R., Salaudeen, Olawale, Chiou, Nicole, Kusner, Matt J., D'Amour, Alexander, Koyejo, Sanmi, Gretton, Arthur
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G1957999595, G2926720356, HDR TRIPODS: Illinois Institute for Data Science and Dynamical Systems (iDS2), NRT-UtB: Training the Next Generation of Researchers in Engineering and Deciphering of Miniature Brain Machinery, CAREER: Probabilistic Models for Spatiotemporal Data with Applications to Dynamic Brain Connectivity, G5620962805, RI: Small: Secure, Private, and Resource-Constrained Approaches to Federated Machine Learning
OpenAlex ID: W4392780962