Research Article
Research on Multiple Classification Based on Improved SVM Algorithm for Balanced Binary Decision Tree
Table 4
The classification accuracy of each multiclassification algorithm fort the dataset “Statlog” in the 10 experiments.
| | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
| OVO | 0.967 | 0.948 | 0.906 | 0.833 | 0.945 | 0.943 | 0.969 | 0.962 | 0.945 | 0.896 | OVA | 0.922 | 0.891 | 0.857 | 0.757 | 0.878 | 0.888 | 0.923 | 0.916 | 0.887 | 0.839 | BDT-SVM | 0.961 | 0.961 | 0.955 | 0.848 | 0.953 | 0.956 | 0.947 | 0.946 | 0.941 | 0.935 | VBDT-SVM | 0.961 | 0.939 | 0.956 | 0.950 | 0.957 | 0.952 | 0.952 | 0.950 | 0.944 | 0.945 | IBDT-SVM | 0.940 | 0.946 | 0.957 | 0.957 | 0.952 | 0.961 | 0.953 | 0.945 | 0.955 | 0.952 |
|
|