Artificial Intelligence for Sustainable Applications

14A Comparison of Selective Machine Learning Algorithms for Anomaly Detection in Wireless Sensor Networks

Arul Jothi S.* and Venkatesan R.

Department of Computer Science and Engineering, PSG College of Technology, Coimbatore, India

Abstract

Wireless Sensor Networks endure from a wide range of faults and anomalies which hinder their smooth working. Anomaly Detection (AD) in wireless sensor networks is a crucial research area, to make sensor nodes to be more protected and consistent. Due to energy constraints and less computation capability of sensor networks AD should focus on the essential boundaries of sensor networks. AD techniques can be categorized as statistical approaches, clustering, and machine learning. Wireless sensor networks observe active environments that vary rapidly overtime. These lively actions are besides the source from peripheral issue or instigated by the developers. In this paper, machine learning techniques that are suitable for anomaly detection and their challenges have been discussed based on the performance metrics factors. Accuracy, precision, and recall contribute to major performance analysis for the implemented model which has been discussed with weather dataset. The research issues of various anomaly detection techniques have been presented with a brief discussion on certain adaptive algorithms.

Keywords: Wireless sensor networks, anomaly detection, machine learning, performance metrics

14.1 Introduction

Sensor networks make use of a device ...