A Gradient Boosting Based Approach for Intrusion Detection in IoT-Enabled Healthcare Systems

Area of Science:

• Cybersecurity
• Medical Informatics
• Machine Learning

Background:

• The integration of Internet of Things (IoT) in healthcare increases cyber-attack risks.
• Existing intrusion detection systems (signature, anomaly) have limitations in recall and adaptability.
• Current machine learning (ML) techniques require further enhancement for emerging cyber-threats.

Purpose of the Study:

• To develop an advanced intrusion detection system (IDS) for IoT-based medical care networks.
• To propose a Gradient Boosting (GB) based technique for improved network security.
• To evaluate the performance of the proposed GB model against other ML algorithms.

Main Methods:

• A Gradient Boosting (GB) based intrusion detection technique was developed.
• The proposed GB model was evaluated using performance metrics: precision, recall, F1-score, and ROC-AUC.
• Comparative analysis was conducted against Random Forest, Extra Trees, Decision Tree, and Naïve Bayes algorithms.

Main Results:

• The Gradient Boosting (GB) technique demonstrated superior performance.
• The proposed model achieved high accuracy and efficiency in detecting intrusions.
• Experimental analysis confirmed the effectiveness of the GB-based approach over existing methods.

Conclusions:

• The GB-based intrusion detection system enhances security in medical care networks.
• The proposed technique offers a more accurate and efficient solution for cyber-attack detection.
• This study highlights the potential of GB for securing sensitive IoT medical environments.