Knowledge-based genetic algorithm for multidimensional data clustering

In this paper, a new approach of genetic algorithm called knowledge-based Genetic Algorithm (KBGA-Clustering) is proposed for multidimensional data clustering. Basically, this method adopts knowledge of what called as appropriate cluster centre for a fixed number of k-cluster. The chromosome which has inappropriate genes will be penalised with maximum value to prohibit it in the next generation. The experimental result is also provided for KBGA-Clustering and Genetic Algorithm-Clustering (GA-Clustering) to present the performance. Based on the observation, KBGA-Clustering presents better performance and more optimum solution compared to conventional GA-Clustering