Association Rules Extraction using Multi-objective Feature of Genetic Algorithm
Mohit Gupta, Geeta Sikka · 2013
Association Rule Mining is one of the most well - liked techniques of data mining strategies whose primary aim is to extract associations among sets of items or products in transactional databases. However, mining association rules typically ends up in a really large amount of found rules, leaving the database analyst with the task to go through all the association rules and find out the interesting ones. Currently Apriori Algorithm plays an important role in deriving frequent itemsets and then extracting association rules out of it. However Apriori Algorithm uses Conjunctive nature of association rules, and single minimum support threshold to get the interesting rules. But these factors don't seem to be alone sufficient to extract interesting association rules effectively. Hence in this paper, we proposed a completely unique approach for optimizing association rules using Multi-objective feature of Genetic Algorithm with multiple quality measures i.e. support, confidence, comprehensibility and interestingness. A global search might be achieved using Genetic Algorithm (GA) in the discovery of association rules as GA relies on the greedy approach. The experiments, performed on numerous datasets, show a wonderful performance of the proposed algorithm and it will effectively reduce the quantity of association rules.