Refining discretizations of continuous-valued attributes

Authors:
Eva Armengol;Àngel García-Cerdaña
Affiliations:
Artificial Intelligence Research Institute (IIIA - CSIC), Campus de la UAB, Bellaterra, Catalonia, Spain;Artificial Intelligence Research Institute (IIIA - CSIC), Campus de la UAB, Bellaterra, Catalonia, Spain,Departament de Lògica, Història i Filosofia de la Ciència, Universitat de Ba ...
Venue:
MDAI'12 Proceedings of the 9th international conference on Modeling Decisions for Artificial Intelligence
Year:
2012

Citing 8
Cited 0

A Distance-Based Attribute Selection Measure for Decision Tree Induction

Machine Learning
Data mining: practical machine learning tools and techniques with Java implementations

Data mining: practical machine learning tools and techniques with Java implementations
Data Mining and Knowledge Discovery Handbook

Data Mining and Knowledge Discovery Handbook
A fuzzy extension of the Rand index and other related indexes for clustering and classification assessment

Pattern Recognition Letters
Clustering and aggregation of relational data with applications to image database categorization

Pattern Recognition
Extending the rand, adjusted rand and jaccard indices to fuzzy partitions

Journal of Intelligent Information Systems
The WEKA data mining software: an update

ACM SIGKDD Explorations Newsletter
Comparing fuzzy, probabilistic, and possibilistic partitions

IEEE Transactions on Fuzzy Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Rand index is a measure commonly used to compare crisp partitions. Campello (2007) and Hüllermeier and Rifqi (2009) respectively, proposed two extensions of this index capable to compare fuzzy partitions. These approaches are useful when continuous values of attributes are discretized using fuzzy sets. In previous works we experimented with these extensions and compared their accuracy with the one of the crisp Rand index. In this paper we propose the ε-procedure, an alternative way to deal with attributes taking continuous values. Accuracy results on some known datasets of the Machine Learning repository using the ε-procedure as crisp discretization method jointly with the crisp Rand index are comparable to the ones given using the crisp Rand index and its fuzzifications with standard crisp and fuzzy discretization methods respectively.