Date of Publication :25th May 2018

Abstract: Most of the real world applications such as Loan approval, Credit card fraud detection etc uses relational databases which contain multiple relations that are inter-linked with the help of primary and foreign keys. It is very tricky to examine these applications with the help of traditional classification methods such as RIPPER and RIDOR. These methods are suitable for single relation and generate simple and comprehensible rules. But it cannot handle uncertainties and noises present in the real dataset. This paper presents a novel method for generating multi-relational classification model for mutagenesis prediction. The classifier is constructed based on the fuzzy extension of the decision tree. The experimental results show the efficiency of the proposed method compared to the existing algorithms

Reference :