Intelligent Identification Apparoch for Healthcare Systems
Rasha Anwar Mohammed, Khattab M. Ali Alheeti · 2022
Intrusion detection is important for healthcare system, where Patient health data is a part of healthcare that is vulnerable to attacks by hackers. These attacks are among the major challenges facing the health system, especially healthcare system which has a weak infrastructure that makes it simpler for these systems to be targeted. Therefore, an accurate attacks detection system must be designed to prevent health data from being manipulated. This paper proposed a system to detect whether there is an attack on healthcare data or not. The system goes through two phases which are data preprocessing & classification. The first phase contains several steps where we merged our existing data, Filling the Missing Values & Scaling. In the second phase, we feed the processed data to three machine learning classifiers, namely Naive Bayes (NB), K-Nearest Neighbors (KNN) & Support Vector Machine (SVM) in order to classify whether there was an attack or not. Through experiments & comparing the results of the three classification algorithms with each other in terms of detection accuracy, the SVM algorithm was found to have the highest classification accuracy rate of 94% compared to KNN algorithm 92% & NB algorithm 90%.