Artificial Intelligence in Human Activity Recognition with Anomaly Prediction for Healthcare systems

Abstract

Understanding human behavior is a complex task that requires careful analysis. Researchers have been exploring various machine learning techniques to gain insights into human nature, but achieving high accuracy has been challenging. In our study, we propose using deep learning techniques to address this issue. To conduct our research, we collected a public dataset using EEG and HRV sensors. To ensure data quality, we applied a median filter to remove noise, specifically salt and pepper noise in the collected images. Next, we employed feature extraction techniques, specifically Genetic Algorithm, to reduce the dimensionality of the datasets and improve accuracy. Finally, we applied classification techniques to further analyze the data. Our proposed approach utilizes an Enhanced Convolutional Neural Network (ECNN) to improve the accuracy of classifications. Finally, we compare our results with the existing approach based on their accuracy. This method helps to detect one's stress level, happiness, and mood swings, among other things.