Table 3 Loss, accuracy and time cost under different batch-size
From: Motor Fault Diagnosis Based on Short-time Fourier Transform and Convolutional Neural Network
Batch-size | Train- loss | Test-loss | Train- acc | Test-acc | Time cost(s) |
|---|---|---|---|---|---|
1 | 0.0597 | 0.000074 | 0.9788 | 1 | 828 |
2 | 0.1105 | 0.000076 | 0.9552 | 1 | 484 |
4 | 0.1369 | 0.000076 | 0.9477 | 1 | 278 |
5 | 0.1585 | 0.000085 | 0.9391 | 1 | 240 |
10 | 0.2468 | 0.0031 | 0.9123 | 1 | 154 |
20 | 0.7756 | 0.0306 | 0.7314 | 0.999 | 123 |
40 | 0.7718 | 0.0456 | 0.7632 | 0.9774 | 77 |
50 | 1.1889 | 0.3048 | 0.59 | 0.9732 | 73 |
80 | 1.3441 | 0.8301 | 0.5319 | 0.7982 | 62 |
100 | 1.5499 | 1.4332 | 0.3985 | 0.5704 | 60 |
200 | 1.5572 | 1.48 | 0.421 | 0.556 | 52 |
400 | 1.5767 | 1.5119 | 0.3995 | 0.6136 | 48 |
500 | 1.593 | 1.5536 | 0.3446 | 0.548 | 50 |