AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Kernel Stein Tests for Multiple Model Comparison

Jen Ning Lim, Makoto Yamada, Bernhard Schölkopf, Wittawat Jitkrittum

NeurIPS 2019 · Learning theory and algorithms › Prediction and decision losses · approach: Distributionally robust optimization · 6 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
Use RelPSI-MMD instead of selecting the lowest FID scoreCelebA image comparison · CelebAmodel selection accuracy: 73%Improved
Use relative testing rather than choosing the model with the minimum FID scoreCelebA image comparison · CelebAidentification of worse candidatesImproved
Use RelPSI-KSD rather than RelMulti-KSDChicago Crime density comparison · Chicago Crimemodel-wise rejection rate: MoG (1): 0.22 -> 0.42; MoG (2): 0.07 -> 0.28; MADE: 0.04 -> 0.26Improved
Use FID minimum-score selection rather than relative MMD testingCelebA image comparison · CelebAincorrect model selection rate: 73%Hurt
Condition the test on the model-selection eventRelPSI versus unconditioned RelMMD/RelKSD · Toy problemstest powerHurt
Use KSD instead of MMDKernel relative tests · Toy problemstest powerImproved
Use linear-time estimators instead of complete estimatorsKernel relative tests · Toy problemstest powerHurt
Use the median-heuristic bandwidth for MMD in the blobs settingMMD-based tests · Blobsrejection rateHurt
Use data splitting for reference selection and testingRelMulti · Multiple model comparisontrue positive rateHurt
Use RelPSI rather than RelMulti under the theorem's stated conditionsTwo-model comparison · Two candidate modelstrue positive rateImproved

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