Table 12 Classification accuracy of different classifiers with human activity recognition using smartphones dataset
From: Selecting critical features for data classification based on machine learning methods
Method | Accuracy | Features |
|---|---|---|
SVM | 0.9796 | 561 |
Boruta + SVM | 0.8692 | 6 |
RFE + SVM | 0.8331 | 6 |
RF + SVM | 0.8685 | 6 |
LDA | 0.9823 | 561 |
Boruta + LDA | 0.7786 | 6 |
RFE + LDA | 0.705 | 6 |
RF + LDA | 0.8297 | 6 |
KNN | 0.9748 | 561 |
Boruta + KNN | 0.864 | 6 |
RFE + KNN | 0.8385 | 6 |
RF + KNN | 0.904 | 6 |
RF | 0.9857 | 561 |
Boruta + RF | 0.8924 | 6 |
RFE + RF | 0.9394 | 6 |
RF + RF | 0.9326 | 6 |