AN ENHANCED OPTIMIZATION APPROACH FOR IMPROVING CLASSIFICATION ACCURACY IN DATA MINING

This work presents an enhanced version of JAYA minimization algorithm to solve the issues in data mining classification approached with different objective functions that reflect in accuracy, reliability, cost, and efficiency improvement. The proposed enhanced JAYA algorithm was obtained by modifying the equations that is used in obtaining the best and worst solutions. Formulation of issues with respect to reliability and efficiency is highlighted in this paper to convert the multi-level objective into mono level objective function. To prove the advantages of proposed EJAYA algorithm two different test sets are used and results are obtained for different scenarios. Conventional genetic algorithm is compared with the obtained results to show the superiority of the proposed EJAYA algorithm with different complexities.