A real time health care cyber attack detection using ensemble classifier

Abstract

Healthcare has been a significant issue for many individuals in recent years. The Internet of things (IoT) and wireless communications may support many medical applications, including early diagnosis and real-time monitoring. Healthcare expenditures may be reduced by using secure and practical strategies to detect life-threatening emergencies in real-time quickly. The aim of the proposed tool is the detection of attacks targeting the Healthcare cyber-physical system. In this method, the wise greedy routing technique is initially used for sensor node placement and creation. The transmitted data is normalized and grouped using an agglomerative mean shift maximization clustering method. The abnormal health features are extracted using a multi-heuristic cyber ant optimization-based feature extraction process. Then the attack is detected by using the Ensemble crossover XG boost classifier. An optimized gradient tree boosting system creates decision trees in a sequential form. From the results, the concluded proposed system effectively monitors patient health conditions, detects data breaches, and improves cloud security. Simulated results show that our proposed strategy improves detection accuracy, increases the proportion of true positives, and reduces false positives by a significant margin.