A Comparative Study of Threat Detection for IoT Devices Using Machine Learning Techniques

“Cybersecurity threats” comprise computer viruses, data breaches, DoS, and other attack methods. These threats try to steal data, corrupt data, or otherwise disrupt digital life. IoT devices lack the required built-in security to combat these types of security threats. As a result of its unsatisfactory hardware and software design and also numerous other factors, IoT devices are at risk of cybersecurity threats. Machine learning techniques are used to handle global computer security issues like malware detection, ransomware recognition, fraud detection, and spoofing identification. This study examines how machine learning is applied to computer defense. Here, the primary focus lies on the prevalent threat of malware (A common threat) and its detection. The fundamental concept used here is that malware is discovered by a machine using the training set after it has been programmed with new data using various algorithms and classification approaches. The most commonly used algorithms for this include supervised neural networks, support vector machine, K-means clustering, KNN (k-nearest neighbors), etc. The results are obtained based on accuracy. Various analyses showed that the highest accuracy was achieved by the SVM algorithm, Decision Tree, and LSTM model. The challenges of malware detection are likewise analyzed.