https://github.com/shimaa83/x-ray

deep learning model to classify x-ray
https://github.com/shimaa83/x-ray

class-weights deep-learning fine-tuning implancec keras vgg16 vgg19

Last synced: about 2 months ago
JSON representation

deep learning model to classify x-ray

• Host: GitHub
• URL: https://github.com/shimaa83/x-ray
• Owner: shimaa83
• Created: 2024-07-31T23:25:34.000Z (6 months ago)
• Default Branch: main
• Last Pushed: 2024-08-04T01:46:24.000Z (6 months ago)
• Last Synced: 2024-12-21T05:42:17.361Z (about 2 months ago)
• Topics: class-weights, deep-learning, fine-tuning, implancec, keras, vgg16, vgg19
• Language: Jupyter Notebook
• Homepage:
• Size: 200 KB
• Stars: 0
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# using deep learning to classify x-ray image

# steps:

1. Import Libraries

2. EDA

3. create dataset

4. images preprocess and use class weight to solve imbalance problem

5. Training the Model

6. Compute Accuracy

7. Making Predictions

8. Conclusion

# we try several model :

1. first model: simple deep learning model 

 :: Accuracy: 0.78 -  Recall: 0.72  - F1 Score: 0.73

2. second model we did some fine tuning by adding drop out layer:

::Accuracy: 0.73  -   Recall: 0.64  - F1 Score: 0.64

3. third model anther fine tuning by adding dense layer:

 ::Accuracy: 0.74 - Recall: 0.66 - F1 Score: 0.65

4. anther pretrained model vgg19

::Accuracy: 0.78  - Recall: 0.70 -F1 Score: 0.72

5. then we try pretrained model vgg16

 ::Accuracy: 0.79   -    Recall: 0.72  - F1 Score: 0.73

 # the highest accuracy and f1score is vgg16 so we use this model in reference