Performed Steps:
Got data from Kaggle public datasets.
Configured TPU (Tensor Processing Unit).
Performed preprocessing and image augmentation on data.
Prepared data for training and validation.
Build a model using transfer learning.
Train and validate the model.
Made predictions.
Project Details:
Python Version: 3.7
Packages: numpy, pandas, kaggle_datasets, tensorflow, keras, matplotlib, efficientnet, sklearn, seaborn, re, math, cv2, PIL, os.
Model: EfficientNet B6 (Transfer Learning with Imagenet weights).
Training Result:
Recall: 0.9542
Precision: 0.9881
Accuracy: 0.9575
Validation Results:
Recall: 0.9744
Precision: 0.9870
Accuracy: 0.9713
Predictions Accuracy: 88.30%
Code Previews:
For full source code please visit the GitHub repository.
Github Repository Link:
0 likes · 11 views · Jul 23, 2024
I created this model to classify pneumonia disease by using the transfer learning technique with the EfficientNet B6 model.
