The authentication of users and devices is essential to the security of cyber-physical systems (CPS). But since various networks and devices are interconnected in CPS, they are vulnerable to cyberattacks, which can have detrimental effects on sectors like healthcare, IoT and blockchain technology. This paper highlights the difficulties faced by CPS in the healthcare system and stresses the value of security and privacy in safeguarding private medical information. The resource limitations, security level specifications, and system architecture of CPS-based healthcare systems, conventional security methodologies and cryptography solutions fall short. In order to better preserve and secure CPS in the healthcare industry, this paper investigates the possibilities of machine learning and multi-attribute feature selection. The suggested solution intends to address the drawbacks of traditional privacy preservation techniques and reduce concerns about sensitive information and data leakage. The security of healthcare data in CPS can be improved by utilizing machine learning techniques, which also aids in the creation of strong network security infrastructures for communication in healthcare applications.words.