ECG-Signal-Classification-for-Arrythmia-Detection

• Built a ECG classifier using MIT-BIH dataset containing ambulatory ECG signals from 47 subjects with 24-hour recordings
• Built on Python and used machine learning libraries such as NumPy, Scikit-learn, Pandas, Matplotlib, Eli5
• Trained the model using Support Vector Machines & Random Forest Classifier with accuracy of ~97% and ~ 98% respectively
• Used Permutation Feature Importance to remove negative features and increased accuracy up to ~2% in each model