Noise Detection in ECG Signal

Overview

Detected noise and artifacts in ECG signals from wearable devices. Applied machine learning techniques for data cleaning, visualization, and feature extraction. Achieved 96.4% accuracy and 91.8% AUPRC using various classification algorithms while addressing class imbalance. Collected real-life data using Shimmer sensors from approximately 6 participants.

Key Highlights

• Detected noise/artifacts in ECG signals from wearable devices
• Applied machine learning techniques for data cleaning, visualization, and feature extraction
• Achieved 96.4% accuracy and 91.8% AUPRC using various classification algorithms while addressing class imbalance
• Collected real-life data using Shimmer sensors from approximately 6 participants

Resources