Built a CNN based deep learning model to solve an image classification problem of detecting Melanoma.
- Built a CNN based model which can accurately detect melanoma. Melanoma is a type of cancer that can be deadly if not detected early. It accounts for 75% of skin cancer deaths. A solution which can evaluate images and alert the dermatologists about the presence of melanoma has the potential to reduce a lot of manual effort needed in diagnosis.
- Creating a basic CNN model, we obserrve that there is significant difference in training accuracy and validation accuracy. This is clear evidence of overfit since the training accuracy is higher than the validation accuracy.
- After using image augmentation and Dropout, the training accuracy has reduced significantly but the validation accuracy has not changed reasonably and the difference betweent the training accuracy and validation accuracy is very minimal. Thus we have reduced overfitting compared to previous model.
- After adding augmented images to manage class imbalance, the training accuracy and the validation accuracy has increased slightly and the difference betweent the training accuracy and validation accuracy is very minimal. Thus we have reduced overfitting compared to the first model and we can get higher accuracy with deeper layers.
This project was based on Melanoma detection by creating a deep learning CNN model.
Created by [@NeerShah87] - feel free to contact me!