Study on data stream prediction based on least square support vector machine

Recently, data stream prediction has become a new hotspot in the field of data processing. Support vector machine is a new forecasting method. It can better solve the small sample, nonlinear, high dimension and local minimum points and other practical problems. In the paper, Sliding window processing mode and incremental online least square support vector machine(LS-SVM) algorithm was introduced to the data flow forecast. The experiment uses the time series data set of gas furnace of Box-Jenkins to validate the effectiveness in the accurate and response time.