Litcius/Paper detail

Looking for Trouble: Analyzing Classifier Behavior via Pattern Divergence

Eliana Pastor, Luca de Alfaro, Elena Baralis

202150 citationsDOI

Abstract

Machine learning models may perform differently on different data subgroups, which we represent as itemsets (i.e., conjunctions of simple predicates). The identification of these critical data subgroups plays an important role in many applications, for example model validation and testing, or evaluation of model fairness. Typically, domain expert help is required to identify relevant (or sensitive) subgroups.

Topics & Concepts

Computer scienceClassifier (UML)Artificial intelligenceMachine learningSimple (philosophy)Identification (biology)Divergence (linguistics)Domain (mathematical analysis)Data miningMathematicsEpistemologyBotanyMathematical analysisPhilosophyLinguisticsBiologyExplainable Artificial Intelligence (XAI)Data Mining Algorithms and ApplicationsMachine Learning and Data Classification