Modified Chaotic Bee Colony Optimization (MCBCO) algorithm for data clustering

Results of the heuristic search-based optimization algorithms largely depend on the initial guess. When the initial guess is closer to the optimal result, then the algorithm converges faster. But for large datasets, the probability of getting this closer guess is difficult. In this paper, a Modified Chaotic Bee Colony Optimization (MCBCO) algorithm is proposed for data clustering. It is capable to explore the solution space in all directions, despite of initial guesses. The chaotic bees that are created using chaotic sequences enable the algorithm to do this. It uses steady state selection tactic for better exploration. The algorithm also uses Gaussian mutation for further exploitations in the solution. The simulation results and analysis reflects that the algorithm is competent for the data clustering problem.