Heartbeat Classification: ML & Deep Learning Pipeline

In the Heartbeat Classification group project, I helped design and optimize an end-to-end modeling pipeline to classify heartbeats from ECG time series. We audited data quality and class imbalance, built strong baselines with classical ML models, and then developed specialized CNN-based architectures that significantly improved performance over these baselines. My focus was on reproducible experimentation, clear evaluation, and structured documentation so that other teams could understand, reproduce, and extend our results.