AI Paper Atlas

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

Tunability: Importance of Hyperparameters of Machine Learning Algorithms

Philipp Probst, Anne-Laure Boulesteix, Bernd Bischl

JMLR 2019 · Learning theory and algorithms › Prediction and decision losses · approach: Active learning and subset selection · 664 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
replace package defaults with data-derived optimal defaultsglmnet · OpenML binary (AUC)tunability: 0.069 -> 0.024Improved
replace package defaults with optimal defaultsrpart · OpenML binary (AUC)tunability: 0.038 -> 0.012Improved
replace package defaults with optimal defaultskknn · OpenML binary (AUC)tunability: 0.031 -> 0.006Improved
replace package defaults with optimal defaultssvm · OpenML binary (AUC)tunability: 0.056 -> 0.042Improved
replace package defaults with optimal defaultsranger · OpenML binary (AUC)tunability: 0.010 -> 0.006Improved
replace package defaults with optimal defaultsxgboost · OpenML binary (AUC)tunability: 0.043 -> 0.014Improved
cross-validate optimal defaultssvm · OpenML binary (accuracy)cross-validated improvement: 0.011 -> -0.001No clear effect
tune two hyperparameters jointly (minbucket and minsplit) instead of the best onerpart · OpenML binary (AUC)joint gain: 0.0055Improved
tune cp jointly with maxdepth instead of cp alonerpart · OpenML binary (AUC)joint gain: 0.0007No clear effect
tune xgboost eta or boosterxgboost · OpenML binary (AUC)tunability: 0.005 and 0.008Improved

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