AI Paper Atlas

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

Harnessing the power of choices in decision tree learning

Guy Blanc, Jane Lange, Chirag Pabbaraju, Colin Sullivan, Li-Yang Tan, Mo Tiwari

NeurIPS 2023 · Learning theory and algorithms › Prediction and decision losses · approach: Optimal decision tree learning · 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
Increase the candidate split count from k to k+1Top-k · Constructed distributionsaccuracy: \frac1{2}+\eps -> 1-\epsImproved
Increase the candidate split count from k to KTop-k · Monotone constructed distributionsaccuracy: \frac{1}{2} + \eps -> 1 - \epsImproved
Increase k across tested settings (ablation)Top-k decision trees · UCI benchmark datasetstest accuracy: > 5\% (absolute)Improved
Increase k beyond the point where test accuracy plateaus (ablation)Top-k decision trees · car and hayes-rothtest accuracyNo clear effect
Increase k after its test-accuracy peak (ablation)Top-k decision trees · tic-tac-toetest accuracyHurt
Use Top-k instead of Top-1 (ablation)Decision trees · taiwanese-bankruptcy, credit-card, and electrical-grid-stabilitytest accuracyHurt
Use Top-k instead of Top-1 (ablation)Decision trees · Most other tested datasetstest accuracyImproved
Increase k from 1 to 16Decision trees · FICO, depth 5, up to 1407 featurestraining scalability: 300 features; 10 minutes -> all 1407 featuresImproved
Increase k from 1 to 16Decision trees · FICO, depth 5, increasing training samplestraining scalability: 150 samples -> 1000 samplesImproved

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