AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

A Unified Theory of Diversity in Ensemble Learning

Danny Wood, Tingting Mu, Andrew M. Webb, Henry W. J. Reeve, Mikel Luján, Gavin Brown

JMLR 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 47 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 split-point randomisation (Random Forest instead of Bagging)decision tree ensemble (m=20) · MNISTgeneralisation error reduction: 9 -> 14.5 (approx. % reduction)Improved
use unconstrained depth trees instead of depth 8 in BaggingBagging, 30 trees · California HousingMSE: 0.35 -> 0.30Improved
use Random Forest instead of unconstrained Bagging30 trees · California HousingMSE: 0.30 -> 0.28Improved
bag 30 trees instead of a single depth-8 treedepth 8 regression tree · California HousingMSE: 0.47 -> 0.35Improved
increase ensemble size m with fixed base learner capacityBagged depth 8 regression trees · California Housingexpected riskImproved
increase tree depth at fixed m=10Bagged regression trees · California Housingbias, variance, diversityMixed
use larger MLPs (100 hidden) instead of small (20 hidden) in BaggingBagged MLPs · MNISTensemble cross-entropyImproved
use Random Forest instead of Bagging with growing ensemble sizeMNIST tree ensembles · MNIST0/1 riskMixed

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