Automatic Weight Calculation Based Associative Classifier

Sapan H Mankad, Swati Gupta

Abstract


We propose a new technique for calculation of weights in weighted associative classification by introducing the concept of Value Frequency (VF) and Inverse Class Frequency (ICF) in classification of relational database. This concept also provides the mechanism for automatic calculation of weights of each item in the relational database for predictions. The VF-ICF is inspired from the famous TF-IDF concept used for document representation in Information Retrieval. It is based on the principle of assigning more importance to a value which is rare in other classes but frequent in a particular class. An initial experiment was carried out to test the proposed approach on Social Workers Dataset and results show that there is a huge scope for improvement.

Keywords


Associative Classification, Weighted Associative Classifier, Value Frequency (VF), Inverse Class Frequency (ICF)

Full Text:

PDF

Refbacks

  • 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