Model Assessment and Selection under Temporal Distribution Shift
This addresses the problem of selecting reliable models in changing environments for machine learning practitioners, though it appears incremental as it builds on existing methods for distribution shift.
The paper tackles model assessment and selection under temporal distribution shift by developing an adaptive rolling window approach to estimate generalization errors and using a tournament for near-optimal model selection, with theoretical and experimental validation of its adaptivity to non-stationary data.
We investigate model assessment and selection in a changing environment, by synthesizing datasets from both the current time period and historical epochs. To tackle unknown and potentially arbitrary temporal distribution shift, we develop an adaptive rolling window approach to estimate the generalization error of a given model. This strategy also facilitates the comparison between any two candidate models by estimating the difference of their generalization errors. We further integrate pairwise comparisons into a single-elimination tournament, achieving near-optimal model selection from a collection of candidates. Theoretical analyses and numerical experiments demonstrate the adaptivity of our proposed methods to the non-stationarity in data.