AI Paper Atlas

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

Robust Data Valuation with Weighted Banzhaf Values

Weida Li, Yaoliang Yu

NeurIPS 2023 · Learning theory and algorithms › Prediction and decision losses · approach: Shapley value attribution · 1 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 best weighted Banzhaf value instead of Banzhafsemi-value · MNISTranking consistency: 0.581 ± 0.075 -> 0.680 ± 0.120Improved
Use best weighted Banzhaf value instead of Banzhafsemi-value · gasranking consistency: 0.729 ± 0.019 -> 0.959 ± 0.009Improved
Use weighted Banzhaf instead of Shapleysemi-value · 2dplanesranking consistency: 0.256 ± 0.083 -> 0.950 ± 0.012Improved
Use weighted Banzhaf instead of Beta Shapley (Beta(16, 1))semi-value · FMNISTranking consistency: 0.017 ± 0.092 -> 0.670 ± 0.042Improved
Use weighted Banzhaf with parameter 0.00 on bioresponsesemi-value · bioresponseranking consistency: 0.948 ± 0.003 -> 0.997 ± 0.000Improved
Use best weighted Banzhaf instead of Banzhaf on lettersemi-value · letterranking consistency: 0.966 ± 0.004No clear effect
Use weighted Banzhaf instead of Banzhafsemi-value · MNIST noisy label detectionF1: 0.625 ± 0.040No clear effect
Use weighted Banzhaf instead of Banzhafsemi-value · 2dplanes noisy label detectionF1: 0.346 ± 0.014 -> 0.609 ± 0.059Improved
Use weighted Banzhaf instead of Shapleysemi-value · cpu noisy label detectionF1: 0.390 ± 0.027 -> 0.772 ± 0.017Improved
Use weighted Banzhaf instead of Shapleysemi-value · bank-marketing noisy label detectionF1: 0.368 ± 0.025 -> 0.528 ± 0.015Improved
Choose small Banzhaf parameter (0.05) for noisy label detectionweighted Banzhaf · most datasetsF1Mixed
Pick the semi-value best at ranking consistency for label detectionsemi-value · all datasetsranking vs F1Hurt

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