AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Benchmarking Distribution Shift in Tabular Data with TableShift

Josh Gardner, Zoran Popovic, Ludwig Schmidt

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Robust fine-tuning for OOD generalization · 12 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Increase in-distribution accuracy across models and tasksTabular classifiers evaluated on TableShift · 15 TableShift tasksout-of-distribution accuracy: ρ=0.81Improved
Apply domain robustness or domain generalization methods instead of standard baselinesTasks with multiple training subdomains · TableShift domain generalization tasksin-distribution accuracyHurt
Apply domain robustness or domain generalization methods instead of standard baselinesTasks with multiple training subdomains · TableShift domain generalization tasksshift gapMixed
Use label-shift robustness methodsTabular classifiers evaluated on TableShift · TableShift tasksshift gapNo clear effect
Use label-shift robustness methodsTabular classifiers evaluated on TableShift · TableShift tasksID and OOD accuracyHurt
Increase change in label distributionModels evaluated across TableShift tasks · TableShift tasksabsolute shift gap: ρ = 0.71Mixed
Use tabular methods beyond standard tree-based baselinesTabular classifiers evaluated on TableShift · 15 TableShift tasksOOD accuracyNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy