Decision Tree: 71%
Random Froest: 83%
SVM: 81% --> very slow at training (585.426215887) comment:yes, and it even takes much longer time to find the optimal penalty parameter updated: SVM 83.5% with Lambda=5 (Run time: 584.36698103)
Logistics: 84%
KNN: 66% --> very slow at testing
Adaboost: 74%
Linear Regressin: has bug comment: I dont see how linear regression could be used for classification ?? We could use it by mapping score >0.5 --> 1 and <0.5 --> 0
Decision Tree: 70%
Random Froest: 82%
SVM:
Logistics: 83%
Logistics with L2 :"testAccuracy": 0.8376196172248804
Logistics with L1: testAccuracy: 0.837320574163
KNN:
Adaboost:
Linear Regressin: has bug comment: I dont see how linear regression could be used for classification ??
#wordDim 2000 --> 4000 dataRate 0.04 The best performance of logistic regression: %85