Bio-inspired algorithm for classification association rules

Associative classification (AC) has shown a great dominance over many classification techniques. Associative classification uses association rule mining for rules discovery process to identify data class labels. Associative classification also integrates the rule discovery and classification process to build the classifier that supports in decision making process. The main advantages of the associative classification approaches is to discover high quality association rules in a very large space of candidate rules and integrate these rules with the classification process efficiently. Artificial Immune Systems (AIS) have emerged during the last decade, Artificial immune systems can be defined as a computational system that is inspired by theoretical immunology, observed immune principles and mechanisms. The AIS uses the population-based search model of evolutionary computation algorithms that it is regarded as a suitable way for dealing with complex search space. This paper proposes an ambitious algorithm based on Quantum-Inspired Immune system (QIS) for building an efficient classifier by searching association rules to find the best subset of rules for all possible association rules. The proposed algorithm employees a mutation operator with a quantum-based rotation gate to control and maintain diversity, and guides the search process. The performance of proposed algorithm is evaluated using benchmark datasets. The experimental results showed that the proposed algorithm is preformed well with large search space and has higher accuracy, and control algorithm diversity.