A Domain Adaptive Feature Relevance Analysis for Human Activity Recognition using Transfer Learning

Wearable electronics are electronic gadgets designed to be incorporated into clothing, worn on the body, or integrated into other items. These devices collect data, provide information, and perform specific tasks. Human Activity Recognition (HAR) utilizes Transfer Learning algorithms to track movements using sensor data from devices like accelerometers and gyroscopes. HAR finds applications in security, fitness tracking, and health monitoring systems. Utilizing pre-trained models and engineered features from these models aids in creating accurate HAR systems. Transfer learning reduces data requirements and simplifies feature engineering, making it a valuable approach for HAR, especially in diverse real-world scenarios.