Discovering Interesting Association Rules: A Multi-objective Genetic Algorithm Approach
Basheer MohamadAl-Maqaleh · International Journal of Applied Information Systems · 2013
Association rule mining is considered as one of the important tasks of data mining intended towards decision making process.It has been mainly developed to identify interesting associations and/or correlation relationships between frequent itemsets in datasets.A multi-objective genetic algorithm approach is proposed in this paper for the discovery of interesting association rules with multiple criteria i.e. support, confidence and simplicity (comprehensibility).With Genetic Algorithm (GA), a global search can be achieved and system automation is developed, because the proposed algorithm could identify interesting association rules from a dataset without having the user-specified thresholds of minimum support and minimum confidence.The experimental results on various types of datasets show the usefulness and effectiveness of the proposed algorithm.