Papers by Keyword: Local Optimum

Abstract: Sadness, one of the negative emotions, may cause undesirable impact to the daily life. Therefore, it is desirable to automatically detect sadness emotion in human-machine interactions in order to adopt measures to impair the negative effects caused by it. Speech is one of the means used by human to express emotions, therefore, it is reasonable to detect sadness emotion using speech samples. In this paper, we analyzed relevant speech features, and proposed an improved Back Propagation (BP) network for sadness recognition. The experimental results show that the improved BP network proposed has better performance than traditional BP networks in detecting sadness emotion.

Abstract: Population diversity is one of the most important factors that influence the convergence speed and evolution efficiency of gene expression programming (GEP) algorithm. In this paper, the population diversity strategy of GEP (GEP-PDS) is presented, inheriting the advantage of superior population producing strategy and various population strategy, to increase population average fitness and decrease generations, to make the population maintain diversification throughout the evolutionary process and avoid “premature” to ensure the convergence ability and evolution efficiency. The simulation experiments show that GEP-PDS can increase the population average fitness by 10% in function finding, and decrease the generations for convergence to the optimal solution by 30% or more compared with other improved GEP.