A review on Discovery of Classification Rules using Pittsburgh Approach



Goal of classification task is to predict the class
which an instance of dataset belongs to. Discovered
knowledge is then presented in the form of high level, easy to
interpret classification rules. Evolutionary algorithms-being
the global search methods- have been widely and
successfully applied for discovery of classification rules from
large datasets. Evolutionary Algorithms have used two
approaches to encode the classification rules: i) Michigan; ii)
Pittsburgh. Michigan approach, initially applied in classifier
systems, does not account for the problem of rule
interaction. Therefore, the Pittsburgh approach seems more
natural for extracting classification rules that evaluates the
whole rule sets and not individual rules. This paper reviews
the encoding schemes, selection strategies, evolutionary
operators, and fitness measures adopted in various learning
classifiers for extracting high level classification rules using
Pittsburgh approach. The review is to provide a firm base to
the researchers who are interested to apply Pittsburgh


Rule mining;pittsburgh approach; review paper; bottom-up approach; classification rule; genetic algorithm; data mining

Full Text:



  • There are currently no refbacks.

Subscribe to Print Journals

 IJAIKD is currently Indexed By   

 http://rgjournals.com/public/site/images/mittalberi/scholar_logo_lg_2011.gif  Journal Seek