AI Paper Atlas

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

On the Epistemic Limits of Personalized Prediction

Lucas Monteiro Paes, Carol Xuan Long, Berk Ustun, Flavio Calmon

NeurIPS 2022 · Learning theory and algorithms › Prediction and decision losses · approach: Data attribution and recourse · 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
add group attributes to a logistic regression (personalization) overalllogistic regression · UCI Adult auditing data, Totalerror rate: 19.5% -> 19.3%Improved
personalize with Sex x Race x Immigration attributeslogistic regression · UCI Adult auditing data, group Male, NW, Rerror rate: 21.6% -> 24.4%Hurt
personalize with Sex x Race x Immigration attributeslogistic regression · UCI Adult auditing data, group Male, W, Rerror rate: 22.1% -> 19.3%Improved
personalize with group attributeslogistic regression · UCI Adult training data, group Male, NW, Rerror rate: 23.0% -> 25.0%Hurt
personalize with group attributeslogistic regression · UCI Adult training data, Totalerror rate: 20.1% -> 20.0%No clear effect
personalize with group attributeslogistic regression · UCI Adult auditing data, group Female, NW, Rgain from personalization: 1.1% -> -1.2%Hurt
increase number of binary group attributes from 18 to 19BoP hypothesis test · n=8x10^9, epsilon=0.01minimax probability of error lower bound: 13% -> 97%Hurt
increase number of samples per group beyond the number of groupsBoP hypothesis test · d=141, epsilon=0.01minimax probability of errorImproved
increase data points per group mBoP estimator · semi-synthetic UCI AdultMSE of BoP estimateImproved
limit dataset to 50k samples with epsilon=0.01BoP hypothesis test · reliable auditingmax binary group attributes: at most 3 binary group attributesHurt

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