关于分类器集成方法的研究及在文本分类中的应用

With more and more methods of classifying based on machine learning，the main research contents is divided into three layers,the first layer is feature extraction ,the second layer is feature selection, the third layer is using the classifier.In statistics and machine learning, ensemble methods use multiple learning algorithms to obtain better predictive performance than could be obtained from any of the constituent learning algorithms.To solve the problem of selecting the optimal ensemble method , a method of counting the features of noise samples to optimize the problem is proposed.The experiment is on the text classification, The results show the validity of this method.