https://github.com/mehrab-kalantari/cnn-chest-xray

Building a Convolutional Neural Network on a smaller sample of chest X-ray dataset
https://github.com/mehrab-kalantari/cnn-chest-xray

class-weights cnn convolutional-neural-networks data-augmentation deep-learning imbalanced-data

Last synced: 4 months ago
JSON representation

Building a Convolutional Neural Network on a smaller sample of chest X-ray dataset

• Host: GitHub
• URL: https://github.com/mehrab-kalantari/cnn-chest-xray
• Owner: Mehrab-Kalantari
• Created: 2023-09-07T22:12:45.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2023-09-07T22:20:03.000Z (almost 2 years ago)
• Last Synced: 2025-01-16T09:42:40.685Z (5 months ago)
• Topics: class-weights, cnn, convolutional-neural-networks, data-augmentation, deep-learning, imbalanced-data
• Language: Jupyter Notebook
• Homepage:
• Size: 164 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# CNN Model on Chest X-ray Dataset

Building a CNN model on chest x-ray dataset. We use a smaller sample of the dataset

Add more data to get a better result

[Dataset on kaggle](https://www.kaggle.com/datasets/paultimothymooney/chest-xray-pneumonia)

## Contents

### Data Augmentation

The data is huge, so we don't load the data directly

There are two classes and the data is imbalanced

### Imbalanced Data Handling

* Class weights

### Modeling

We use the following architecture for our model

![p](sample/model.png)

### Model Improving

* Early stopping

* ReduceLROnPlateau

### Evaluation

* Accuracy and loss plotting