IMPROVEMENT OF IOT SECURITY WITH A MACHINE LEARNING BASED INTRUSION DETECTION SYSTEM APPROACH

Abstrak

The development of the Internet of Things (IoT) has brought convenience to various
aspects of life, but it also presents significant challenges regarding cybersecurity. One solution
to address this issue is the development of an Intrusion Detection System (IDS) based on machine
learning. This study aims to design an efficient and adaptive IDS for IoT environments using
machine learning algorithms such as Random Forest and Support Vector Machine (SVM). The
methodology includes system design, data collection, algorithm selection, model training, and
system performance evaluation. The results show that Random Forest and SVM algorithms are
effective in detecting attacks such as Distributed Denial of Service (DDoS) and malware, with a
relatively high accuracy rate. However, the main challenges faced are the need for representative
datasets and computational efficiency issues on resource-constrained IoT devices. This study
concludes that machine learning-based intrusion detection systems can improve IoT security by
accurately detecting cyber-attacks. Further development is expected to address efficiency
constraints and enhance the system's reliability in facing increasingly complex threats.