DECISION TREE BASED ON HIPPOPOTAMUS OPTIMIZATION ALGORITHM FOR SECURING IOT HEALTHCARE SYSTEMS

This paper presents how the Internet of Things (IoT) has transformed healthcare by improving system efficiency and reducing the need for human intervention, while highlighting the security vulnerabilities arising from its interconnected nature. Intrusion Detection Systems (IDS) play a critical role in mitigating these threats, and recent advancements in machine learning and artificial intelligence have significantly enhanced detection capabilities. Among the various techniques, decision tree-based models have proven particularly effective in handling the large and complex data flows typical of IoT environments. To further improve security, the study incorporates advanced encryption methods and proposes the Hippopotamus Optimization Algorithm (HOA), which simulates the behavior of hippopotamuses to optimize IDS performance. By fine-tuning decision trees and other classifiers, HOA achieves higher classification accuracy and more efficient anomaly detection. Comparative analysis of machine learning algorithms—such as Logistic Regression, Decision Trees, K-Nearest Neighbors (KNN), Support Vector Machines (SVM), K-Means clustering, and Random Forest—demonstrates that ensemble and deep learning models are more robust for securing IoT healthcare systems. Although models like Random Forest and KNN show high detection accuracy, challenges such as class imbalance remain. The proposed HOA-based hybrid model addresses these limitations by optimizing the precision-recall trade-off, ultimately providing a more resilient security framework for IoT-based healthcare applications.