Generalization bounds can be obtained deterministically via sensitivity analysis of optimization problems, with probabilistic assumptions used ex post to bound the error term measuring closeness of out-of-sample to in-sample data.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Separating Geometry from Probability in the Analysis of Generalization
Generalization bounds can be obtained deterministically via sensitivity analysis of optimization problems, with probabilistic assumptions used ex post to bound the error term measuring closeness of out-of-sample to in-sample data.