AI Paper Atlas

Research map / Learning theory and algorithms / Submodular and game algorithms

Optimal Pricing for Data-Augmented AutoML Marketplaces

Minbiao Han, Steven Xia, Jonathan Light, Raul Castro Fernandez, Haifeng Xu, sainyam galhotra

ICML 2026 · Learning theory and algorithms › Submodular and game algorithms · approach: Learned auction design

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
search over external datasets with bandit-based joint augmentation-model discovery instead of AutoML-only model searchData-Bandit · NYC Open Data classification tasksaverage utility: around 0.55 -> higher than 0.85Improved
use bandit-based model selection instead of training all models each iteration (Data-All) or fixed RF (Data-Alt)Data-Bandit · NYC Open Data classification tasksaverage utility: just above 0.7 -> higher than 0.85Improved
use MILP-optimized price curve instead of Independent, Shift, Jiggle pricingpricing mechanism · School data, in-sampleprofit fraction of welfare: less than 80% -> around 85%Improved
use MILP pricing out of sample with restricted metric poolpricing mechanism · School data, out-of-sampleprofit fraction of welfareImproved
increase learning rate for prior learning (1/(t+1) to 1/sqrt(t) to 1/2) (ablation)Bayesian prior learning · Random prior distributionconvergence/stabilityMixed
add batch averaging of posteriors (larger batch size) (ablation)Bayesian prior learning · Random prior distributionstability of KL divergenceImproved

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