Table 3. Average running time of different feature selection techniques.
| Feature selection technique | Feature selection time | Classification time |
| Correlation filter | 4.00 | 4.86 |
| 1-HC- TOP | 9.71 | 4.71 |
| 5×10−2-HC-TCOMP | 27.00 | 3.00 |
| 5×10−2-HC-TOP | 26.29 | 5.43 |
| 5×10−3-HC-TCOMP | 27.57 | 2.43 |
| 5×10−3-HC-TOP | 27.86 | 4.43 |
| 1-HIGH-TOP | 1.43 | 4.29 |
| 5×10−2-HIGH-TCOMP | 4.86 | 1.71 |
| 5×10−2-HIGH-TOP | 4.86 | 4.43 |
| 5×10−3-HIGH-TCOMP | 4.57 | 1.71 |
| 5×10−3-HIGH-TOP | 4.43 | 4.71 |
Times are in seconds per iteration of the leave-one-out cross validation. The SVM-radial and KNN classifiers were used after the feature selection phase. These tests were done using a Linux platform with Intel Xeon 5160 CPU, 3.00 GHz clock, and 4 GB of RAM.