AI Paper Atlas

Research map / Graph learning / Graph algorithms and clustering

On the Discrimination Risk of Mean Aggregation Feature Imputation in Graphs

Arjun Subramonian, Kai-Wei Chang, Yizhou Sun

NeurIPS 2022 · Graph learning › Graph algorithms and clustering · approach: Graph distance and retrieval · 0 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 0-fair imputation instead of regular Global MeanGlobal Mean imputation · SBMdiscrimination risk: 0.051 -> 0Improved
use 0-fair imputation instead of regular Neighbor MeanNeighbor Mean imputation · SBMdiscrimination risk: 0.048 -> 0Improved
use eps-fair constraint with eps=0 instead of regular on Feature PropagationFeature Propagation · SBMdiscrimination risk: 0.028 -> 0Improved
use eps-fair constraint with eps=0 on Graph RegularizationGraph Regularization · SBMdiscrimination risk: 0.021 -> 0Improved
0-fair instead of regular Global MeanGlobal Mean imputation · SBMrelative reconstruction error: 1 -> 1.21Hurt
0-fair instead of regular Feature PropagationFeature Propagation · SBMrelative reconstruction error: 0.977 -> 1.184Hurt
0-fair instead of regular Graph RegularizationGraph Regularization · SBMrelative reconstruction error: 0.977 -> 1.006Hurt
0-fair Feature Propagation instead of regular, linear group-membership classifierFeature Propagation · SBMgroup membership identification accuracy (linear): 0.814 -> 0.6Improved
0-fair instead of regular, GCN group-membership classifierFeature Propagation · SBMgroup membership identification accuracy (gcn): 0.612 -> 0.505Improved
0-fair instead of regular, MLP group-membership classifierGraph Regularization · SBMgroup membership identification accuracy (mlp): 0.821 -> 0.713Improved
lower eps from 0.05 to 0 in fair Global Mean (ablation)eps-fair Global Mean · SBMgroup membership identification accuracy (linear): 0.736 -> 0.602Improved
eps-fair constraints on Feature Propagation / Graph Regularization (structure-aware) vs fair GM (ablation)eps-fair imputation · SBMrelative reconstruction error: 1.21 -> 1.006Improved

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