Table 4. Computational Adjustments for AlexNet, GoogLeNet, and ResNet50 Training Parameters.
| MODEL | OBS | TS
(OBS x 70%) |
VS
(OBS x 30%) |
MBS | IPE
(TS/MBS) |
EPOCHS | MI
(EPOCH X IPE) |
|---|---|---|---|---|---|---|---|
| AlexNet | 9860 | 6902 ≈ (9860 x 0.7) | 2958 ≈ (9860 x 0.3) | 138 | 50 ≈ (6902/138) | 30 | 1500 = (30 x 50) |
| ResNet50 | 50 | 138 ≈ (6902/50) | 30 | 4140 = (30 x 138) | |||
| GoogLeNet | 138 | 50 ≈ (6902/138) | 30 | 1500 = (30 x 50) |
Glossary: Observations (OBS), Test Set (TS), Validation Set (VS), Mini Batch Size (MBS), Iteration Per Epoch (IPE), Maximum Iterations (MI).