A Text Clustering Algorithm based on Weeds and Differential Optimization

Abstract

Invasive weed optimization (IWO) is a swarm optimization algorithm with both explorative and exploitive power where the diverisity of the population is obtained by allowing the reproduction and mutation of individuals with poor fitness .Differential optimization algorithm is a random parallel algorithm according to a vector change that can make individuals change toward outstanding individuals with global convergence.For k-means algorithm , the traditional algorirhm is prone to get stuck at local optimum and is sensitive to random initialization. Based on the aforementiond background a novel optimization algorithm based hybriding DE and IWO which denoted IWODE-KM is employed to optimize the parameters of k-means and is further applied to chinese text clustering. Experiment results shows that the proposed method outperforms both of its ancestors.