AI Paper Atlas

Research map / Learning theory and algorithms / Fairness

Certification of Distributional Individual Fairness

Matthew Robert Wicker, Vihari Piratla, Adrian Weller

NeurIPS 2023 · Learning theory and algorithms › Fairness · approach: Fairness-constrained classification

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 F-IBP local IF regularization instead of FTU2-layer MLP · Folktables IncomeA-DFC: 1.000 -> 0.130Improved
train with F-IBP local IF regularization instead of FTU2-layer MLP · Folktables Incomeaccuracy: 0.820 -> 0.771Hurt
train with SenSR instead of FTU2-layer MLP · Folktables EmployA-DFC: 0.999 -> 0.639Improved
train with SenSR instead of FTU2-layer MLP · Folktables Employaccuracy: 0.809 -> 0.743Hurt
add L-DIF distributional regularization on top of F-IBP2-layer MLP · Folktables CoverageA-DFC: 0.288 -> 0.136Improved
add L-DIF distributional regularization on top of F-IBP2-layer MLP · Folktables Coverageaccuracy: 0.699 -> 0.698No clear effect
use U-DIF upper-bound regularization instead of L-DIF2-layer MLP · Folktables IncomeA-DFC: 0.095 -> 0.042Improved
use U-DIF upper-bound regularization instead of L-DIF2-layer MLP · Folktables Incomeaccuracy: 0.763 -> 0.725Hurt
use U-DIF instead of L-DIF2-layer MLP · Folktables EmployE-DFC: 0.051 -> 0.006Improved
increase network size (number of neurons)L-DIF certification · DIF upper boundupper bound valueHurt
increase gamma from 0.01 to 0.2DIF upper/lower bounds · 2x16 NNbound gap: 0.01 -> 0.2Hurt
train with L-DIF instead of MILP trainingsingle hidden layer NN · training timetraining time: 4096 neurons under one hourImproved

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