Intrusion Detection in Smart Healthcare Using Bagging Ensemble Classifier

Abstract

Critical infrastructure systems such as ehealthcare, which are crucial to government, society, and commerce, are directly related to people's lives. In order to offer extremely secured critical infrastructure systems, a reliable, scalable, and robust threat detection system must be established to effectively alleviate cyber threats. Moreover, big data from threat monitoring systems cause severe trials for cyber-attack since increasing number of devices in the structure and the quantity of monitoring data accumulated from IoT based healthcare devices need to prevent cyberattacks. Intrusion detection systems, or IDSs, have become an important component in maintaining the security on a corporate network. The popularity of Internet usage comprises many risks of network attacks. Intrusion detection aims to discover unusual access or attacks to secure internal networks. To build our proposed intrusion detection system, Machine learning techniques are utilized with benchmark attack data. In this paper, Bagging ensemble classifier for intrusion detection in a smart healthcare environment is proposed. Bagging ensemble classifier with Random Forest achieved better performance (97.67% accuracy) as compared to its counterpart classifier models for intrusion detection in a smart healthcare with a benchmark data. Experimental results have revealed the practicality of Bagging ensemble classifier by accomplishing a better performance for the cyber security threats in a smart healthcare environment.