Table 2 Class-wise data partitioning for model training and testing.
Class | Original Samples | Synthetic AI Samples | Total Samples |
|---|---|---|---|
Normal track | 720 | 400 | 3,360 |
Superelevation | 720 | 400 | |
Wheel burn | 720 | 400 |
Train Test Split | Training Samples | Testing Samples | Total Samples |
|---|---|---|---|
Normal track | 896 | 224 | 3,360 |
Superelevation | 896 | 224 | |
Wheel burn | 896 | 224 |
