ECG Arrhythmia Detection with Deep Learning

This project classifies ECG heartbeats into arrhythmia types using a 1D Convolutional Neural Network (CNN) trained on the MIT-BIH dataset.

Features

• Full pipeline: data loading, preprocessing, splitting, and scaling
• CNN model for time-series ECG signal classification
• Training/validation accuracy and loss curves
• Model evaluation: confusion matrix, classification report, and accuracy score

Model Accuracy vs Validation

Stack Used

• Python (NumPy, pandas, matplotlib, seaborn)
• scikit-learn (data preprocessing, metrics)
• TensorFlow/Keras (model architecture, training)
• Jupyter Notebook