Detecting Anomalies in IoT Using Intuitionistic Fuzzy Clustering Algorithms

Abstract

Intuitionistic Fuzzy Sets (IFS) are convenient ways to represent the vagueness and uncertainty inherent in any dataset. There are many uses of these. One such use is the Fuzzy C-Means (FCM) algorithm, which clusters the data objects into a pre-assigned (c) number of fuzzy clusters, which has been adopted in many fields, namely, pattern classification, anomaly detection, fraud identification, etc. One of the important applications is finding anomalies in Internet of Things (IoT) data. IoT is made up of a vast network of digital devices that are constantly producing enormous amounts of data and performing live calculations. Owing to their high susceptibility to the Internet, IoT nodes frequently face issues from illegitimate access, such as intrusions, anomalies, and fraud. Detecting such anomalies in the IoT domain might be a fascinating research challenge. In this article, we propose to develop and evaluate Intuitionistic Fuzzy Clustering (IFC) and Interval-Valued Intuitionistic Fuzzy Clustering (IVIFC) methods for the detection of IoT anomalies by extending the widely acknowledged FCM algorithm and establishing their efficacies through complexity analysis, experimental studies, and comparative analysis.