Table 2. Performance indicators of the CNN and other algorithms for the three individual classes.
Class/Indicators | Insertion | Withdrawal | Stop | |
---|---|---|---|---|
CNN | Recall | 0.958 | 0.946 | 0.967 |
Precision | 0.976 | 0.969 | 0.918 | |
F1 score | 0.967 | 0.958 | 0.942 | |
VGG-16 | Recall | 0.95 | 0.916 | 0.964 |
Precision | 0.953 | 0.959 | 0.911 | |
F1 score | 0.952 | 0.937 | 0.937 | |
LeNET | Recall | 0.907 | 0.833 | 0.872 |
Precision | 0.874 | 0.909 | 0.83 | |
F1 score | 0.89 | 0.87 | 0.85 | |
SVM | Recall | 0.844 | 0.775 | 0.799 |
Precision | 0.804 | 0.832 | 0.785 | |
F1 score | 0.823 | 0.802 | 0.792 |