A Hybrid Optimization Approach for Human Activity Using Smartphone Sensors

In the digital age, smartphones equipped with advanced sensors offer opportunities for predicting human activities. This paper presents a hybrid model, combining support vector machines (SVM) and random forests (RF), along with a soft voting classifier, to accurately predict various daily activities in real-time using smartphone inertial sensors. Our approach includes dataset acquisition, preprocessing, and feature extraction, demonstrating robust performance across different activities. The result section represents the performance of our hybrid model in comparison to benchmark schemes.