Certified Learning under Distribution Shift: Sound Verification and Identifiable Structure

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gokavarapu, Chandrasekhar, Gadde, Sudhakar, Rajasekhar, Y., Bhargava, S. R.
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915808629751808
author Gokavarapu, Chandrasekhar
Gadde, Sudhakar
Rajasekhar, Y.
Bhargava, S. R.
author_facet Gokavarapu, Chandrasekhar
Gadde, Sudhakar
Rajasekhar, Y.
Bhargava, S. R.
contents Proposition. Let $f$ be a predictor trained on a distribution $P$ and evaluated on a shifted distribution $Q$. Under verifiable regularity and complexity constraints, the excess risk under shift admits an explicit upper bound determined by a computable shift metric and model parameters. We develop a unified framework in which (i) risk under distribution shift is certified by explicit inequalities, (ii) verification of learned models is sound for nontrivial sizes, and (iii) interpretability is enforced through identifiability conditions rather than post hoc explanations. All claims are stated with explicit assumptions. Failure modes are isolated. Non-certifiable regimes are characterized.
format Preprint
id arxiv_https___arxiv_org_abs_2602_17699
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Certified Learning under Distribution Shift: Sound Verification and Identifiable Structure
Gokavarapu, Chandrasekhar
Gadde, Sudhakar
Rajasekhar, Y.
Bhargava, S. R.
Machine Learning
Rings and Algebras
68T05, 62G35, 62G20, 49J20, 90C26
Proposition. Let $f$ be a predictor trained on a distribution $P$ and evaluated on a shifted distribution $Q$. Under verifiable regularity and complexity constraints, the excess risk under shift admits an explicit upper bound determined by a computable shift metric and model parameters. We develop a unified framework in which (i) risk under distribution shift is certified by explicit inequalities, (ii) verification of learned models is sound for nontrivial sizes, and (iii) interpretability is enforced through identifiability conditions rather than post hoc explanations. All claims are stated with explicit assumptions. Failure modes are isolated. Non-certifiable regimes are characterized.
title Certified Learning under Distribution Shift: Sound Verification and Identifiable Structure
topic Machine Learning
Rings and Algebras
68T05, 62G35, 62G20, 49J20, 90C26
url https://arxiv.org/abs/2602.17699