AI Paper Atlas

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

Being Properly Improper

Tyler Sypherd, Richard Nock, Lalitha Sankar

ICML 2022 · Learning theory and algorithms › Prediction and decision losses · approach: Ranking-based loss optimization · 2 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
Replace log-loss (alpha=1) boosting with alpha-loss (alpha in {1.1,2,4}) via PM-BoostPM-Boost on xd6 with feature noise · xd6 feature noise p=0.5accuracy: 0.96 vs 0.88 AdaBoost, 0.83 XGBoost at p=0.5Improved
Larger alpha (1.1 to 4.0) in PM-Boost (ablation)PM-Boost on xd6 feature noise · xd6 feature noise p=0.5accuracy: 0.91 at alpha=1.1 and 2.0 to 0.96 at alpha=4.0Improved
Alpha-loss boosting vs AdaBoost/XGBoost under no or low feature noisePM-Boost on xd6 · xd6 feature noise p=0 to 0.25accuracy: all methods at or near 1.00No clear effect
PM-Boost with alpha-loss on noisy class labels (symmetric label noise) versus proper log-loss boostingPM-Boost on cancer dataset with class noise · cancer class noise 0/15/30%accuracy: qualitative gains over AdaBoost and XGBoost when twist presentImproved
Adaptive alpha (Menon PM-Boost, alpha estimated from data) versus fixed alpha=1.1 (ablation)PM-Boost on xd6 with class noise · xd6 class noiseaccuracy: slightly better in the paper's wordsMixed
Focal loss generalization of log-lossfocal loss · theorytwist-propernessHurt
Super Loss wrapperSuper Loss · theorytwist-propernessHurt

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