The study on the result prediction and comparison of College English Test Band 4 in China based on Support Vector Machine

摘要

This paper applies Support Vector Machine (SVM) algorithm to carry on Chinese College English Test Band 4 (CET-4) forecast research, forms the sample set with 2000- 2009 year CET-4 data, establishes SVM model including the influencing factor and CET-4 result. Use SVM on the training and learning for input and output data, approach functional relations which the historical data conceals, complete mapping of new data series, thus complete the future year CET-4 result prediction and carry on the comparison with several other methods' forecast effect. The result indicates that the precision of the SVM model surpasses other forecast method.