AI Paper Atlas

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

Complete Search for Feature Selection in Decision Trees

Salvatore Ruggieri

JMLR 2019 · Learning theory and algorithms › Prediction and decision losses · approach: Optimal decision tree learning · 18 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 black-box sequential backward elimination with DTsbe (enumeration of distinct trees)SBE · 20 UCI datasets (m=2)elapsed time ratio: speedup in the range 2x - 100xImproved
use DTsbe instead of SBE on high-dimensional datasetsSBE · Madelon, Isolet, Gisette, P53mutants, Arcene, Dextercompletes within 1h time-out: SBE >1h vs DTsbe 9.79 s on MadelonImproved
use DTdistinct to avoid building duplicate treesSBE enumeration · benchmark datasetsduplicate trees built: at most 0.035% duplicate treesImproved
add heuristic feature selection (DTsbe) to decision treesdecision tree on all features · Adult, 10-fold CVgeneralization error: 15.76 -> 14.45Improved
add complete search (DTaccept0, best subsets) instead of DTsbeDTsbe · Adultgeneralization error: 14.45 -> 14.30Improved
use complete search (DTaccept0) instead of heuristic DTsbeDTsbe · Ionosphere and Annealgeneralization error: Ionosphere 9.77 -> 11.29Hurt
oversearching with complete search (ablation)feature selection · Sonargap between estimated and generalization error: DTaccept0 estimated 2.12 vs generalization 26.40Hurt
add DTsbe feature selection before random forestRF · Dextergeneralization error: 48.95 -> 43.59Improved
add DTsbe feature selection before random forestRF · Madelongeneralization error: 36.25 -> 20.93Improved
add DTsbe feature selection before random forestRF · Sonargeneralization error: 18.42 -> 23.26Hurt
use random forest instead of single decision tree with feature selectionDTsbe · 20 UCI datasetsgeneralization error: RF worse only on Hypo and MadelonImproved
use all instances for training instead of splitting into build and search sets (ablation)feature selection · Ionosphere, large mgeneralization error: baseline best, then SBE, then best subsetsImproved

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