Comparison of Information Theoretical Measures for Reduct Finding
نویسندگان
چکیده
The paper discusses the properties of an attribute selection criterion for building rough set reducts based on discernibility matrix and compares it with Entropy and Gini index used for building decision trees. It has been shown theoretically and experimentally that Entropy and Gini index work better if the reduct is later used for prediction of previously unseen cases, and the criterion based on the discernibility matrix tends to work better for learning functional relationships where generalization is not an issue.
منابع مشابه
Attribute Reduction and Information Granularity
This work was supported by Science and Technology Commission of Shanghai Municipality, No.705931 ABSTRACT In the view of granularity, this paper analyzes the influence of three attribute reducts on an information system, finding that the possible reduct and − μ decision reduct will make the granule view coarser, while discernible reduct will not change the granule view. In addition, we investig...
متن کاملA Heuristic Optimal Reduct Algorithm
Reduct finding, especially optimal reduct finding, similar to feature selection problem, is a crucial task in rough set applications to data mining, In this paper, we propose a heuristic reduct finding algorithm, which is based on frequencies of attributes appeared in discernibility matrix. Our method does not guarantee to find optimal reduct, but experiment shows that in most situations it doe...
متن کاملDimension Reduction Using Rough Set Theory For Intrusion Detection System
This paper briefly describes Information Detection System management by reducing dimension of the information system(IS) for network traffic both row wise and column wise. For dimension reduction, Rough Set Theory (RST) is used where determination of reduct and core plays a major role. But finding reduct is difficult task for a huge IS. Our paper suggests a technique where we propose to divide ...
متن کاملA General Definition of an Attribute Reduct
A reduct is a subset of attributes that are jointly sufficient and individually necessary for preserving a particular property of a given information table. A general definition of an attribute reduct is presented. Specifically, we discuss the following issues: First, there are a variety of properties that can be observed in an information table. Second, the preservation of a certain property b...
متن کاملInformation Quantity-based Decision Rule Acquisition from Decision Tables
Decision rule acquisition is widely used in data mining and machine learning. In this paper, the limitations of the current approaches to reduct for evaluating decision ability are analyzed deeply. Two concepts, i.e. information entropy and information quantity, and the process of constructing decision tree for acquiring decision rule are introduced. Then, the standard of classical significance...
متن کامل