Getting the Best Rules using Genetic Algorithm Variants

A large amount of transaction data gets stored on storage devices every day. To convert this huge amount of data into useful information we have proposed some data mining techniques which represent datasets (with the help of rules) to a maximum extent. A popular method to generate such rules is Association Rule Mining. The popular rule mining algorithm named Apriori algorithm can be used to generate these rules. However the efficiency and optimized set of rules can be achieved using Genetic algorithms and its variants. The present paper elaborates on a system developed to find all the possible optimized rules from the data set generated by Apriori algorithm using Genetic Algorithm and using its variants to improve the performance of the system by speeding up the time required for getting the results after scanning huge datasets. Experimentally a speedup of 1.67 using Parallel Genetic Algorithms and a speedup of 1.76 using Hierarchical Genetic Algorithms respectively as compared to using Simple Genetic Algorithms is generated by the system proposed.