LLAD: Life-Log Anomaly Detection Based on Recurrent Neural Network LSTM.

Area of Science:

• Biomedical Engineering
• Health Informatics
• Artificial Intelligence in Healthcare

Background:

• Daily health monitoring, termed Life-Log, utilizes sensor data to identify health anomalies.
• Smart sensors enhance personal healthcare by increasing awareness of health conditions and wellness.
• Effective Life-Log analysis is crucial for real-time health monitoring and anomaly detection.

Purpose of the Study:

• To propose an improved and combined method for detecting anomalies in sensor-generated health log data.
• To enhance the accuracy and effectiveness of health data analysis for personal healthcare.

Main Methods:

• Utilizing recurrent neural networks (RNNs) with long short-term memory (LSTM) units for Life-Log data analysis.
• Integrating and improving upon existing techniques for anomaly detection in health data.

Main Results:

• The proposed model demonstrates superior performance compared to conventional health data analysis methods.
• The approach achieves satisfactory accuracy in identifying anomalies within health log data.

Conclusions:

• The developed method offers a more effective solution for anomaly detection in personal healthcare.
• This approach enhances the capabilities of smart sensors for continuous health monitoring and risk mitigation.