The aim of this paper is to show how abduction can be used in classification tasks when we deal with incomplete data. Some classifiers, even if based on decision tree induction like C4.5 [1], produce as output a set of rules in order to classify new given examples. Most of these rule-based classifiers make the assumption that at classification time we can know all about new given examples. Probabilistic approaches make rule-based classifiers able to get the most probable class, on the basis of the frequency of the missing attribute in the training set [2]. This kind of assumption sometimes leads to wrong classifications. We present an abductive approach to (help to) choose which classification rule to apply when a new example with missing information needs to be classified, using knowledge about the domain.

Abduction in Classification Tasks

MANCARELLA, PAOLO MARIA;TURINI, FRANCO
2003

Abstract

The aim of this paper is to show how abduction can be used in classification tasks when we deal with incomplete data. Some classifiers, even if based on decision tree induction like C4.5 [1], produce as output a set of rules in order to classify new given examples. Most of these rule-based classifiers make the assumption that at classification time we can know all about new given examples. Probabilistic approaches make rule-based classifiers able to get the most probable class, on the basis of the frequency of the missing attribute in the training set [2]. This kind of assumption sometimes leads to wrong classifications. We present an abductive approach to (help to) choose which classification rule to apply when a new example with missing information needs to be classified, using knowledge about the domain.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

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: http://hdl.handle.net/11568/201647
 Attenzione

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

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 1
social impact