An improved genetic approach

In this paper, we propose an improved genetic algorithm, which is based on an incremental genetic K-means algorithm. This approach combines an incremental genetic algorithm with K-means clustering. The main difference of our approach from the original lies in that we get rid of illegal solutions, which were allowed in the original, during whole evolution process of the genetic algorithm from initialization to its termination. The improvement in our approach is accomplished through changing the way of generating initial population in initialization phase and changing the method of dealing with empty clusters in mutation operation. Thus, the illegal solutions were eliminated from our algorithm and resulting more efficient time performance. Experimental results show that our improved genetic approach is promising