Mangalagiri Nanda Gopal

I built a privacy-preserving ECG-based heart disease classification system using federated learning. It trained Scikit-learn classifiers on ECG signal features and achieved 93% accuracy without centralizing patient data.

Using the Flower framework, I simulated federated training across distributed clients. The project used ECG datasets including PTB Diagnostic ECG Database, MIT-BIH Arrhythmia Database, PTB-XL, and the Chapman-Shaoxing 12-lead ECG dataset.

I also built a Python pipeline to clean historical temperature and rainfall data and visualize seasonal and long-term trends. I’m a final-year B.Tech student in Artificial Intelligence & Data Science, looking to apply my skills in a data analyst or machine learning trainee role.