An Empirical Comparison of Machine Learning Algorithms for Attack Detection in Internet of Things Edge

Abstract

This research work is aimed to perform a comparative analysis of different machine learning algorithms for attack detection at the Internet of Things (IoT) edge. Due to the rising development of IoT, attack detection has become extremely important in network security, as it protects the IoT network from suspicious activities. The self-configuring and open nature of IoT devices is vulnerable to both internal and external attacks. The statistical method of attack detection is not suitable for fast and accurate detection due to the multi-dimensional nature of attacks. Machine learning-based edge computing can rectify these issues through automated response and shifting the computation physically closer to the device edge where the information is generated. In this paper, we have compared the performances of eight machine learning (ML) algorithms to identify the optimal ML algorithm for attack detection in IoT Edge.

DL-ADS: Improved Grey Wolf Optimization Enabled AE-LSTM Technique for Efficient Network Anomaly Detection in Internet of Thing Edge Computing

“…Researchers have used various machine learning (ML) and DL algorithms in recent years to increase the performance of an IDS. Decision Tree (DT) [8], Self-Organizing Map (SOM) [9], K-Nearest Neighbors (KNN) [8], Logistic Regression (LR) [8], Support Vector Machine (SVM) [8], and Random Forest (RF) [8] are the most frequently used ML algorithms for anomaly detection. Nonetheless, applying conventional ML algorithms to large, noisy, and complex platforms is challenging since they primarily rely on manually extracted features and lack labeled training datasets.…”

IGWO-SoE: Improved Grey Wolf Optimization Based Stack of Ensemble Learning Algorithm for Anomaly Detection in Internet of Things Edge Computing

“…In previous research, individual ML algorithms have been examined, but their false alarm rates and detection rates are not more effective. Many ML algorithms have been considered for building IDS, such as decision tree (DT) [11], dimensionality reduction algorithms [12], random forest (RF) [13], swarm intelligence techniques [14], support vector machine (SVM) [15], K-nearest neighbor (KNN) [16], logistic regression (LR) [16], and naive Bayes (NB) [17]. However, designing a robust anomaly detection model using a single ML algorithm is a challenging endeavor.…”