AI Paper Atlas

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

Bridging the Gap: Unifying the Training and Evaluation of Neural Network Binary Classifiers

Nathan Tsoi, Kate Candon, Deyuan Li, Yofti Milkessa, Marynel Vazquez

NeurIPS 2022 · Learning theory and algorithms › Prediction and decision losses · approach: Loss landscape analysis · 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
train with soft-set F1 loss instead of BCEneural binary classifier · AdultF1-Score: 0.26 ± 0.06 -> 0.63 ± 0.02Improved
train with soft-set F1 loss instead of BCEneural binary classifier · MammographyF1-Score: 0.56 ± 0.11 -> 0.63 ± 0.04Improved
train with soft-set F1 loss instead of BCEneural binary classifier · KaggleF1-Score: 0.50 ± 0.33 -> 0.83 ± 0.02Improved
train with soft-set F1 loss instead of BCEneural binary classifier · CocktailPartyF1-Score: 0.70 ± 0.02 -> 0.75 ± 0.01Improved
soft-set F1 loss instead of adversarial F1 approachneural binary classifier · CocktailPartyF1-Score: 0.30 ± 0.06 -> 0.75 ± 0.01Improved
optimize Accuracy with soft sets instead of BCEneural binary classifier · CocktailPartyAccuracy: 0.85 ± 0.01 -> 0.85 ± 0.01No clear effect
optimize Accuracy with soft sets on imbalanced dataneural binary classifier · MammographyF1-Score: 0.56 ± 0.11 -> 0.00 ± 0.00Hurt
optimize AUROC with soft sets instead of BCEneural binary classifier · CocktailPartyAUROC: 0.78 ± 0.01 -> 0.57 ± 0.00Hurt
optimize AUROC via WMW statistic instead of BCEneural binary classifier · AdultAUROC: 0.58 ± 0.02 -> 0.50 ± 0.00Hurt
train with soft-set F1 loss instead of BCECNN · CIFAR-10-FrogF1-Score: 0.59 ± 0.04 -> 0.73 ± 0.01Improved
train with soft-set Accuracy loss instead of BCECNN · CIFAR-10-TransportationF1-Score: 0.88 ± 0.01 -> 0.92 ± 0.00Improved
increase beta in F-beta loss from 1 to 3neural binary classifier · MammographyRecall: 0.55 ± 0.07 -> 0.81 ± 0.03Improved

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