We propose a novel rule-based method that explains the prediction of any classifier on a specific instance by analyzing the joint effect of feature subsets on the classifier prediction. The relevant subsets are identified by learning a local rule-based model in the neighborhood of the prediction to explain. While local rules give a qualitative insight of the local behavior, their relevance is quantified by using the concept of prediction difference

Enhancing Interpretability of Black Box Models by means of Local Rules / Pastor, Eliana; Baralis, ELENA MARIA. - ELETTRONICO. - (2019). ((Intervento presentato al convegno 6th ACM Celebration of Women in Computing: womENcourage 2019 tenutosi a Rome (Italy) nel 16-18 September 2019.

Enhancing Interpretability of Black Box Models by means of Local Rules

Eliana Pastor;Elena Baralis
2019

Abstract

We propose a novel rule-based method that explains the prediction of any classifier on a specific instance by analyzing the joint effect of feature subsets on the classifier prediction. The relevant subsets are identified by learning a local rule-based model in the neighborhood of the prediction to explain. While local rules give a qualitative insight of the local behavior, their relevance is quantified by using the concept of prediction difference
File in questo prodotto:
File Dimensione Formato  
womENcourage_2019_paper_69.pdf

accesso aperto

Descrizione: Abstract
Tipologia: Abstract
Licenza: PUBBLICO - Tutti i diritti riservati
Dimensione 280.29 kB
Formato Adobe PDF
280.29 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2752953
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo