Table 13.
Early integration results for S2783 data set.
k-NN | DT | SVM | SVR | |||||
cv | test | cv | test | cv | test | cv | test | |
SO | 0.795 | 0.794 | 0.748 | 0.762 | 0.829 | 0.832 | 0.825 | 0.828 |
SO* | 0.793 | 0.794 | 0.751 | 0.756 | 0.829 | 0.829 | 0.824 | 0.827 |
TO | 0.804 | 0.803 | 0.762 | 0.769 | 0.821 | 0.824 | 0.813 | 0.818 |
TO* | 0.806 | 0.799 | 0.770 | 0.780 | 0.826 | 0.829 | 0.818 | 0.824 |
ST | 0.797 | 0.797 | 0.758 | 0.766 | 0.829 | 0.831 | 0.825 | 0.828 |
ST* | 0.798 | 0.797 | 0.766 | 0.782 | 0.829 | 0.830 | 0.825 | 0.828 |
The accuracy of each base-learner trained with original data and with extra features added in SO/SO*, TO/TO* or ST/ST*. The values reported for each classifier and regressor are respectively the validation and test accuracies of the original representation and the new representation.