A time series prediction algorithm by support vector machine and application to grain yield prediction

We present a new time series prediction algorithm based on support vector machine (SVM) and linear auto-regression (AR) theory; at the same time, a new method of determining time series embedding dimension for SVM time series prediction is designed. Furthermore, the China grain yield prediction models are formulated. The experimental results on China grain yield data from publicly available datasets indicate that the proposed model for SVM time series prediction is effective and has an obvious advantage of generalization compared with the traditional method.