Chaotic time series prediction based on fuzzy possibility c-mean and composite kernel support vector regression

A clustering based composite kernels support vector machine ensemble forecasting model is proposed for the chaotic time series prediction. First, fuzzy possibility c-mean clustering algorithm (FPCM) is adopted to partition the input dataset into several subsets, which can overcome the drawback caused by outlier and noise in conventional fuzzy c-mean method. Then, SVMs with composite kernels that best fit partitioned subsets are constructed respectively, which hyperparameters are adaptively evolved by immune clone selection algorithm (ICGA). Finally, a fuzzy synthesis algorithm is employed to combine the outputs of submodels to obtain the final output, in which the degrees of memberships are generated by the relationship between a new input sample data and each subset center. Simulation results on a chaotic benchmark time series indicate that the presented algorithm shows good prediction performance compared to the other existing algorithms for the time series prediction task considered in this paper.