An analysis of metaheuristic algorithms for optimization with data clustering in data mining for classification

A metaheuristic provides a sufficiently good solution to an optimization problem. In this paper, we propose few metaheuristic Algorithms. In this paper, we give the pros and cons of PSO, ACO, and BA algorithm. We show the superiority of the new metaheuristic bat algorithm (BA) over other standard algorithms such as ACO and Particle Swarm Optimization. For comparison purpose, we have studied and analyzed few Swarm Intelligence (SI) algorithms on different datasets. The experiments are conducted on attributes which are categorical and continuous for the Fi classification. attribute are continuous. The performance of every classifier was evaluated by calculating the weighted Arithmetic mean, Normalized Absolute error, Standard Deviation and accuracy. The experimental results show that the BAT classifier performs better than all the compared algorithms.