Table 3 Hardware, software, and model training configuration.
From: Integrating transfer learning with scalogram analysis for blood pressure estimation from PPG signals
Item | Content | Model description | Parameters of random forest | ||
|---|---|---|---|---|---|
Pretrained model | Output dimension | ||||
CPU | 12th Gen Intel(R) Core(TM) i5-12500H 2.50 GHz | VGG16 | 7 × 7 × 512 | No. of Estimators | 100 |
GPU | NVIDIA GEFORCE RTX 3050 | ResNet50 | 7 × 7 × 2048 | Random State set during Training process | 42 |
RAM | 8.00 GB | InceptionV3 | 5 × 5 × 2048 | Validation Method | Tenfold cross validation |
Operating System | Windows 11 | NASNetLarge | 11 × 11 × 4032 | No. of Epochs | 30 |
Programming Language | Python 3.6 | InceptionResNetV2 | 8 × 8 × 1536 | Performance Metrics | MAE, SD |
Environment Used | Jupyter Notebook | ConvNeXtTiny | 7 × 7 × 768 | Accuracy percentage when considered error difference of < 5, < 10, < 15 mmHg | |
Deep learning Library | TensorFlow 2.10.0 | Input to Random Forest (The output of pretrained models are flattened and fed input to Random Forest) | 25,088 × 1 100,352 × 1 51,200 × 1 487,872 × 1 98,304 × 1 37,632 × 1 | ||
