Empowering Cyber-Physical Systems through AI-driven Fusion for Enhanced Health Assessment

Abstract

Cyber-physical systems (CPS) for improved health assessment currently face multiple challenges with managing and treating health issues, and there is an uncontrollable amount of risk, as well as the requirement for effective artificial intelligence (AI) algorithms that can manage complex and dynamic health data. In this paper, we provide Hybrid Support Vector Machines fine-tuned Spatial Transformer Networks (HSVM+FSTN) for the prediction of enhanced health assessment. 299 cardiac failure patients were gathered in the Kaggle source, and the data was pre-processed using Z-score normalization. Linear discriminant analysis was employed for feature extraction. Classification was provided for both the machine learning (ML) and deep learning (DL) techniques. The performance analysis was carried out on the Python platform. The proposed HSVM+FSTN performance was more significant when compared to existing techniques in terms of sensitivity (95.4%), accuracy (97.2%), specificity (94.4%), and precision (96.7%).