AI Paper Atlas

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

Beyond Trees: Classification with Sparse Pairwise Dependencies

Yaniv Tenzer, Amit Moscovich, Mary Frances Dorn, Boaz Nadler, Clifford Spiegelman

JMLR 2020 · Learning theory and algorithms › Prediction and decision losses · approach: Tree ensemble optimization · 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 HSIC-based filtering of near-independent bivariate densities (ablation)SLB · General BN, Gaussian, unbalanced, n=1000misclassification error (%): 5.29 -> 4.11Improved
add HSIC-based filtering of near-independent bivariate densities (ablation)SLB · General BN, Gaussian, balanced, n=200misclassification error (%): 12.3 -> 11.7Improved
add HSIC-based filtering of near-independent bivariate densities (ablation)SLB · General BN, Gaussian, balanced, n=1000misclassification error (%): 3.73 -> 3.67No clear effect
add HSIC-based filtering of near-independent bivariate densities (ablation)SLB · General BN, Gaussian, unbalanced, n=200misclassification error (%): 15.1 -> 15.1No clear effect
add bivariate log-density features to univariate onesSLB vs LU · Ionospherebalanced error rate (%): 45.4 -> 7.5Improved
add bivariate log-density features to univariate onesSLB vs LU · Pimabalanced error rate (%): 49.2 -> 28.6Improved
replace tree-structured TAN with SLB under general BNsSLB · General BN, Gaussian, balanced, n=1000misclassification error (%): 10.8 -> 3.67Improved
replace TAN with SLB on forest-structured Gaussian dataSLB · Forest BN, Gaussian, balanced, n=200misclassification error (%): 8.17 -> 15.4Hurt
increase training sample size from 200 to 1000SLB · Forest BN, Gaussian, balancedmisclassification error (%): 15.4 -> 5.11Improved
replace SLB with random forest on complex forest dataSLB · Forest BN, complex, balanced, n=1000misclassification error (%): 27.6 -> 23.8Hurt
replace TAN with SLB on real dataSLB · Ionospherebalanced error rate (%): 7.6 -> 7.5No clear effect
replace NB with SLB on OzoneSLB · Ozonebalanced error rate (%): 24.9 -> 32.3Hurt

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