利用python进行特征选择 通过比较每一个特征对于整体分类的贡献, 计算DXscore,从而进行特征的排序, 对于每一个排序进行选择,从最强的特征开始组合, 最终我们确定最佳的方案组合 在最佳的方案组合下,我们利用网格搜索对SVM搜索参数, 最终我们得到最佳的C和gamma 利用最佳的参数再次进行搜索 得到最终的强特
-
Notifications
You must be signed in to change notification settings - Fork 0
imaginistLi/feature_select
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
About
利用python进行特征选择
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published