Incremental and Decremental Algorithms of Fuzzy Support Vector Regressor and Its Application

摘要

A fuzzy support vector regressor (FSVR) modeling method and its incremental and decremental algorithms are proposed in this paper. Based on which, an on-line FSVR modeling method based on sliding time window is also proposed. Which uses the samples in the time window to build the dynamic system model, and with the slide of the time window, the proposed incremental and decremental algorithms are used to update the trained FSVR without from scratch. The proposed method is applied in predicting the yield of acrylonitrile, study results demonstrate the effectiveness of this method.