Core Reduct Based Preprocessing Approach to Incomplete Data

Abstract

Most of the data mining algorithm's application hampers due to missing attribute values. Inadequate treatment of missing values seriously affects the data mining and classification accuracy. A useful technique has been proposed to deal with missing attribute values. Rough set approach to incomplete information system has been shown. Application of discernible matrix for incomplete information to compute core and reduct has been shown. Imputation based preprocessing approach depends on relation between present attribute value and incomplete attribute value, so have to found most similar object to impute missing value. To find similar objet importance or priority should be given in core attributes, after that reduct attributes if matching occur in corresponding attributes value and other attributes may be neglected. In this paper this concept of core and reduct attributes of rough set has been utilized to fill missing values using the proposed Core Reduct Based (CRB) algorithm. Efficiency of the CRB algorithm in the completeness analysis to incomplete data has been shown by comparing its performance with other existing algorithms using some real life data sets.

Attribute Reduction With Imputation Of Missing Data Using Fuzzy-Rougsh Set

“…Rough set can deal with missing values efficiently. Concepts of indiscernibility relation and discernibility matrix of rough set is very popular to fill missing values [1]- [2]. Rough sets core and reduct concepts used to impute missing value using most similar object [1]- [2].…”

“…Concepts of indiscernibility relation and discernibility matrix of rough set is very popular to fill missing values [1]- [2]. Rough sets core and reduct concepts used to impute missing value using most similar object [1]- [2]. New concept tolerance relation of rough set introduced by kryszkiewicz and for incomplete information dispensability of attributes, indispensability of attributes, core, and functional dependency has been redefined for rough set attributes [12].…”

“…Most similar object used for imputation. For most similar objet core attributes have highest priority and reduct attributes are also considered if matching occurs in corresponding attribute values [1]- [2]. So using FRARI algorithm complete reduce data set can be generated for efficient and effective data mining.…”