AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Multiclass Anomaly Detector: the CS++ Support Vector Machine

Alistair Shilton, Sutharshan Rajasegarar, Marimuthu Palaniswami

JMLR 2020 · Label-efficient and robust learning › OOD and anomaly detection · approach: Anomaly detection with contaminated data

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 CS++-SVM single-step anomaly detection instead of hybrid 1-class SVM plus CS-SVMartificial 5-class data · Artificial 5-class data, class 2correct classification rate (%): 59.6% -> 90.41%Improved
use CS++-SVM instead of hybrid, effect on false anomaly labels for class 2artificial 5-class data · Artificial 5-class data, class 2labelled anomalous (%): 36.1% -> 4.1%Improved
use CS++-SVM instead of hybrid on digit versus character data (RBF kernel)DIG/CHR character recognition · CHR anomalies versus DIGanomaly labellingImproved
increase nu to large values (ablation)CS++-SVM · DIG/CHR RBF kernelclassification and anomaly detection accuracyHurt
use CS++-SVM instead of hybrid on HARHAR · HARnon-anomaly score NA%: 90.8% -> 96.4%Improved
use CS++-SVM instead of hybrid on DSADSA · DSAnon-anomaly score NA%: 69.0% -> 90.3%Improved
use CS++-SVM instead of hybrid on ForestForest · Forestnon-anomaly score NA%: 1.4% -> 87.8%Improved
use CS++-SVM instead of classify-with-reject on HAR anomaly classHAR · HARanomaly class score A%: 1.0% -> 99.5%Improved
use CS++-SVM instead of classify-with-reject on DSA non-anomaly scoreDSA · DSAnon-anomaly score NA%: 96.7% -> 90.3%Hurt

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