Evaluation results for MALLET ten-fold cross validation with leave-one-out feature.
收藏数据链接：
官方服务：
资源简介：
This overview shows the individual importance of each of the features in the overall classification model. The large majority of features have very little impact over the model, i.e., a decrease in performance of 1–2%. The only two features that make a difference are the Prefix and the token context (Token_Bi3) that affect the overall performance with almost 15%.
创建时间：
2015-12-02
