International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 379 ISSN 2229-5518 IJSER © 2014 http://www.ijser.org An Evolutionary Quantum Behaved Particle Swarm Optimization for Mining Association Rules K. Indira1, S. Kanmani2, R. Jagan3, G. Balaji3, F. Milton Joseph3 1Research Scholar, Department of Computer Science and Engineering 2Professor, Department of Information Technology 3Final Year UG Student, Department of Information Technology Pondicherry Engineering College, Puducherry, India. induharini@gmail.com1, kanmani@pec.edu2, jaganit@pec.edu3, bala@pec.edu3, itsmemilton@gmail.com3 Abstract- In data mining, association rule mining is a popular and well researched method for discovering interesting relations between variables in large databases, which are meaningful to the users and can generate strong rules on the basis of these frequent patterns, which are helpful in decision support system. Quantum Particle Swarm Optimization (QPSO) is one of the several methods for mining association rules. It combines the aspects of traditional PSO philosophy and quantum mechanics. However, preventing the occurrence of local optima and improving the convergence speed is still a tedious task. In this paper, an Evolutionary Quantum behaved Particle Swarm Optimization (EQPSO) is presented with improved computational efficiency and has proper convergence. The proposed work introduces local search techniques into QPSO using Modified Shuffled Frog Leaping Algorithm (MSFLA) and depicts a systematic parameter adaptation by developing an Evolutionary State Estimation (ESE) and an Elitist Learning Strategy (ELS). The EQPSO implementation has comprehensively been evaluated on 5 different datasets taken up from the UCI Irvine repository. The performance of EQPSO is compared with Basic QPSO and the experimental results shows that the proposed system outperforms the existing algorithm quite significantly. Keywords- Association Rule Mining, Elitist, Evolutionary State, Memetic, Particle Swarm Optimization, Quantum Behavior, Self-Adaptive, Shuffled Frog Leaping. ——————————  —————————— 1 INTRODUCTION Data Mining, the analysis step of Knowledge Discovery in Databases, is the practice of examining large pre-existing databases in order to generate new information. However, continuous increase in the amount of data stored in databases and types of databases, the extraction of critical hidden information from these databases has become tedious. Several methods such as classification, clustering and association rules have been used to deduce interferences from large databases. Among these methods, association rule mining is the most widely used method. Association rules (AR) are usually required to satisfy a user-specified minimum support and a user-specified minimum confidence at the same time [1]. The association rule generation can be split up into steps: firstly, to find all frequent itemsets in a database, minimum support value is applied. Second, the rules are formed making use of these frequent itemsets and the minimum confidence constraints. Apriori algorithm is one of the first known association rule mining algorithm. It uses a level wise search, where k-itemsets are used to explore (k+1)-itemsets, to mine frequent itemsets from transactional database for Boolean association rules. In order to improve the efficiency, a major step forward was the introduction of compact data structure, referred to FP-tree or frequent pattern tree and its association mining algorithm FP-growth. However, support calculation is possible only if the entire dataset is added to the tree. Then, an alternative of Apriori Itemset Generation called Dynamic Itemset Counting was introduced. In this algorithm, once the transactions are read, the itemsets are dynamically added and deleted. It relies on the fact that for an itemset, all of its subsets must also be frequent, so only those itemsets whose subsets are all frequent can be examined. Later, evolutionary algorithms like Genetic Algorithm (GA) [2], Particle Swarm Optimization (PSO) [3] and a variant of PSO namely Quantum Particle Swarm Optimization (QPSO) were introduced. These algorithm are population based search methods and it moves from one set of points (population) to another set of points in a single iteration with likely improvement using set of control operators. However, the fundamental problem in the existing system is its imbalanced local and global search and IJSER