Human-like rule optimization for continuous domains

When using machine learning techniques for data mining purposes one of the main requirements is that the learned rule set is represented in a comprehensible form. Simpler rules are preferred as they are expected to perform better on unseen data. At the same time the rules should be specific enough s...

Full description

Bibliographic Details
Main Authors: Hadzic, Fedja, Dillon, Tharam S.
Other Authors: Ana Fred
Format: Book Chapter
Published: Springer 2008
Subjects:
Online Access:http://hdl.handle.net/20.500.11937/42310
_version_ 1848756384816955392
author Hadzic, Fedja
Dillon, Tharam S.
author2 Ana Fred
author_facet Ana Fred
Hadzic, Fedja
Dillon, Tharam S.
author_sort Hadzic, Fedja
building Curtin Institutional Repository
collection Online Access
description When using machine learning techniques for data mining purposes one of the main requirements is that the learned rule set is represented in a comprehensible form. Simpler rules are preferred as they are expected to perform better on unseen data. At the same time the rules should be specific enough so that the misclassification rate is kept to a minimum. In this paper we present a rule optimizing technique motivated by the psychological studies of human concept learning. The technique allows for reasoning to happen at both higher levels of abstraction and lower level of detail in order to optimize the rule set. Information stored at the higher level allows for optimizing processes such as rule splitting, merging and deleting, while the information stored at the lower level allows for determining the attribute relevance for a particular rule. The attributes detected as irrelevant can be removed and the ones previously detected as irrelevant can be reintroduced if necessary. The method is evaluated on the rules extracted from publicly available real world datasets using different classifiers, and the results demonstrate the effectiveness of the presented rule optimizing technique.
first_indexed 2025-11-14T09:11:21Z
format Book Chapter
id curtin-20.500.11937-42310
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T09:11:21Z
publishDate 2008
publisher Springer
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-423102022-12-07T06:50:50Z Human-like rule optimization for continuous domains Hadzic, Fedja Dillon, Tharam S. Ana Fred Joaquim Filipe Hugo Gamboa Feature Selection Data Mining Rule Optimization When using machine learning techniques for data mining purposes one of the main requirements is that the learned rule set is represented in a comprehensible form. Simpler rules are preferred as they are expected to perform better on unseen data. At the same time the rules should be specific enough so that the misclassification rate is kept to a minimum. In this paper we present a rule optimizing technique motivated by the psychological studies of human concept learning. The technique allows for reasoning to happen at both higher levels of abstraction and lower level of detail in order to optimize the rule set. Information stored at the higher level allows for optimizing processes such as rule splitting, merging and deleting, while the information stored at the lower level allows for determining the attribute relevance for a particular rule. The attributes detected as irrelevant can be removed and the ones previously detected as irrelevant can be reintroduced if necessary. The method is evaluated on the rules extracted from publicly available real world datasets using different classifiers, and the results demonstrate the effectiveness of the presented rule optimizing technique. 2008 Book Chapter http://hdl.handle.net/20.500.11937/42310 10.1007/978-3-540-92219-3_25 Springer fulltext
spellingShingle Feature Selection
Data Mining
Rule Optimization
Hadzic, Fedja
Dillon, Tharam S.
Human-like rule optimization for continuous domains
title Human-like rule optimization for continuous domains
title_full Human-like rule optimization for continuous domains
title_fullStr Human-like rule optimization for continuous domains
title_full_unstemmed Human-like rule optimization for continuous domains
title_short Human-like rule optimization for continuous domains
title_sort human-like rule optimization for continuous domains
topic Feature Selection
Data Mining
Rule Optimization
url http://hdl.handle.net/20.500.11937/42310