Timezone: »

 
Poster
An Information-theoretic Approach to Distribution Shifts
Marco Federici · Ryota Tomioka · Patrick Forré

Fri Dec 10 08:30 AM -- 10:00 AM (PST) @
in Poster Session 8 »

Safely deploying machine learning models to the real world is often a challenging process. For example, models trained with data obtained from a specific geographic location tend to fail when queried with data obtained elsewhere, agents trained in a simulation can struggle to adapt when deployed in the real world or novel environments, and neural networks that are fit to a subset of the population might carry some selection bias into their decision process.In this work, we describe the problem of data shift from an information-theoretic perspective by (i) identifying and describing the different sources of error, (ii) comparing some of the most promising objectives explored in the recent domain generalization and fair classification literature. From our theoretical analysis and empirical evaluation, we conclude that the model selection procedure needs to be guided by careful considerations regarding the observed data, the factors used for correction, and the structure of the data-generating process.

Keywords: Theory · Deep Learning · Machine Learning · Graph Learning · Domain Adaptation · Representation Learning

Author Information

Marco Federici (University of Amsterdam)
Ryota Tomioka (Microsoft Research Cambridge)
Patrick Forré (University of Amsterdam)

More from the Same Authors