Implementing ensemble of deep learning model with optimization techniques for human activity recognition to assist individuals with disabilities

Recent human activity recognition (HAR) developments have allowed numerous applications like healthcare, smart homes, and improved manufacturing. Activity recognition plays a crucial part in improving human well-being by capturing behavioral data, enabling computing systems to analyze, monitor, and assist the daily lives of individuals with disabilities. HAR assist in comprehending complex human actions and behaviors by observing body movements and the surrounding environment, aiming to support individuals with disabilities, and is therefore extensively utilized in smart homes, athletic competitions, medical care, and other applications. This paper proposes a binary grey wolf optimization-driven ensemble deep learning model for human activity recognition (BGWO-EDLMHAR) technique. The BGWO-EDLMHAR technique aims to develop a robust HAR method for disability assistance using advanced optimization models. The data normalization stage applies z-score normalization at first to transform raw data into a clean and structured format suitable for analysis or modelling. Furthermore, the binary grey wolf optimization (BGWO) method is utilized for the feature selection process. Moreover, the ensemble of DL models, namely bidirectional long short-term memory, variational autoencoder, and temporal convolutional network methods, are employed for the classification process. Finally, the cetacean optimization algorithm optimally adjusts the ensemble models' hyperparameter values, resulting in more excellent classification performance. The BGWO-EDLMHAR approach is examined using the WISDM dataset. The comparison study of the BGWO-EDLMHAR approach portrayed a superior accuracy value of 98.51% over existing models.