This project was completed for CS230: Deep Learning as taught by Stanford University. The goal was to diagnose pneumonia through the use of X-ray images, which were fed to a convolutional neural network.
The approach was to use a modified version of ResNet-50 (the only modification was to add Dropout layers for exploratory purposes). The dataset was taken from Kaggle [1].
Several ideas were explored in this project, including:
- Dropout Layers
- Learning Rate Decay
- Transfer Learning
- Regularization
- Data Augmentation
- Combining augmented and normal data generators
The model was able to achieve over 96% accuracy, thus proving automated pneumonia detection through deep learning is efficient.
[1] https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia