AI Paper Atlas

Research map / Learning theory and algorithms / Clustering

Consistency, Breakdown Robustness, and Algorithms for Robust Improper Maximum Likelihood Clustering

Pietro Coretto, Christian Hennig

JMLR 2017 · Learning theory and algorithms › Clustering · approach: Clustering approximation algorithms

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
use RIMLE with an improper noise density instead of plain Gaussian mixture (mclust)clustering · AsyNoise (n=500, p=20, 37.4% noise)misclassification rate: 61.4% -> 8.8%Improved
use RIMLE instead of mclust with noise componentclustering · AsyNoise (n=500, p=20, 37.4% noise)misclassification rate: 38% -> 8.8%Improved
use RIMLE instead of TCLUST with true trimming rateclustering · AsyNoise (n=500, p=20, 37.4% noise)misclassification rate: 49% -> 8.8%Improved
use RIMLE instead of best-tuned DBSCANclustering · AsyNoise (n=500, p=20, 37.4% noise)misclassification rate: 26% -> 8.8%Improved
use RIMLE instead of mclust without noise componentclustering · GEM (n=100, p=20, 2% outliers)misclassification rate: 2% -> 100% accurateMixed
use RIMLE instead of mclust with noise componentclustering · GEM (n=100, p=20, 2% outliers)misclassification rate: 47% -> 100% accurateImproved
use RIMLE instead of TCLUST with 2% trimmingclustering · GEM (n=100, p=20, 2% outliers)misclassification rate: 11% -> 100% accurateImproved
choose log(delta) below -100 versus within [-53,-36]RIMLE · AsyNoise (n=500, p=20, 37.4% noise)misclassification rate: values below -100 do not change resultsNo clear effect
increase log(delta) to very large valuesRIMLE · AsyNoise (n=500, p=20, 37.4% noise)estimated noise proportion: noise proportion gets close to zeroHurt
select delta with OTRIMLE criterionRIMLE · AsyNoise (n=500, p=20, 37.4% noise)selected log(delta): log(delta) = -40Improved

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