A Survey Of Machine Learning Techniques For Detecting Anomaly In Internet Of Things (IoT)

• Published in: Journal of independent studies and research computing Vol. 21; no. 1
• Main Authors: Imran, Imran, Ali, Syed Mubashir, Faiz, Rizwan, Alam, Muhammad Mansoor, Quadri, Syed Kashif Ali, Bari, Muhammad Razeen, Farooq Shibli, Muhammad
• Format: Journal Article
• Language: English
• Published: Shaheed Zulfikar Ali Bhutto Institute of Science and Technology 07.06.2023
• ISSN: 2412-0448; 1998-4154
• DOI: 10.31645/JISRC.23.21.1.5

In recent years, there has been a lot of focus on anomaly detection. Technological advancements, such as the Internet of Things (IoT), are rapidly being acknowledged as critical means for data streams that create massive amounts of data in real time from a variety of applications. Analyzing this gathered data to detect abnormal occurrences helps decrease functional hazards and avoid unnoticed errors that cause programme delay. Methods for evaluating specific anomalous behaviorsin IoT data stream sources have been established and developed in the current literature. Unfortunately, there are very few thorough researches that include all elements of IoT data acquisition. As a result, this article seeks to address this void by presenting a comprehensive picture of numerous cutting-edge solutions on the fundamental concerns and essential issues in IoT data. The data type, types of anomalies,the learning method, datasets, and evaluation criteria are all described. Lastly, the issues that necessitate further investigation and future approaches are highlighted.