Support vector regression based on optimal training subset and adaptive particle swarm optimization algorithm

Figure optionsDownload as PowerPoint slideHighlights
• A parameters selection algorithm of SVR is proposed based on OTS.
• The APSO algorithm has been combined to determine the size of OTS.
• The proposed model has the best accuracy and generalization ability by experiments.
• The optimal parameters setting of SVR and the optimal size of OTS are studied.
• The obtained results confirm the applicability and superiority of the proposed model.

Support vector regression (SVR) has become very promising and popular in the field of machine learning due to its attractive features and profound empirical performance for small sample, nonlinearity and high dimensional data application. However, most existing support vector regression learning algorithms are limited to the parameters selection and slow learning for large sample. This paper considers an adaptive particle swarm optimization (APSO) algorithm for the parameters selection of support vector regression model. In order to accelerate its training process while keeping high accurate forecasting in each parameters selection step of APSO iteration, an optimal training subset (OTS) method is carried out to choose the representation data points of the full training data set. Furthermore, the optimal parameters setting of SVR and the optimal size of OTS are studied preliminary. Experimental results of an UCI data set and electric load forecasting in New South Wales show that the proposed model is effective and produces better generalization performance.