🧠 GAN Comparative Study: DCGAN vs Conditional GAN in Low-Resource Settings

A deep dive into the performance comparison of DCGAN and Conditional GAN under resource-constrained environments, designed especially for healthcare applications.

---

📌 Overview

This project focuses on comparing two powerful generative models — DCGAN and Conditional GAN — in low-resource environments. The models are evaluated using multiple metrics and visualization techniques such as:

• Loss Curves (Generator and Discriminator)
• Generated Image Quality Comparison
• Performance Radar Charts (Accuracy, Precision, Recall, F1-Score, AUC)
• Heatmap of Feature Correlation
• PCA and t-SNE for Feature Reduction and Cluster Visualization

Application Domain: Healthcare – specifically for generating synthetic patient records and medical images where real data is limited.

---

🧪 Visual Results

1. Generated Images

2. Loss Curves Comparison

3. Heatmap Comparison

4. Performance Radar Chart

5. PCA and t-SNE Visualizations

---

🚀 How to Run Locally

git clone https://github.com/yourusername/GAN-Comparison-Healthcare.git
cd GAN-Comparison-Healthcare
pip install -r requirements.txt
python main.py

---

🌱 How to Contribute

1. Fork this repository (click here).
2. Create your branch: git checkout -b feature-name
3. Make your changes and commit: git commit -m "Add feature"
4. Push to your forked repo: git push origin feature-name
5. Submit a pull request 🚀

---

💡 Key Technologies

• Python - Core implementation
• TensorFlow - GAN model creation
• Matplotlib & Seaborn - Visualizations
• PCA / t-SNE - Dimensionality Reduction

---

📬 Contact

Developed by: Sunil R
Email: sunilr31r@gmail.com
GitHub: yourusername

---

📄 License

This project is licensed under the MIT License - feel free to use, modify, and share.