Machine Learning and Big Data Analytics for Cyber Security Intrusion Detection

The growing prominence of the Internet of Things (IoT), fast development of network computers, and the proliferation of apps have all contributed to an elevated focus on cyber security. Identifying different cyber-attack or irregularities in a network, while also developing an effective detection structure, has grown increasingly important in modern security. A data-driven systems for intrusion detection can be created utilizing intelligent machines, notably neural networks. The paper presents a Mayfly optimization (MFO) with regularized extreme learning machine (RELM) approach, known as the MFO-RELM computer-learning-based security system, which prioritizes security aspects and creates a generalized intrusion detecting model based on the given variables. This approach correctly predicts previously unseen situations in tests while reducing computational complexity by reducing feature size. We evaluated our MFO-RELM algorithm’s effectiveness using cyber security databases, calculating precision, recall, F-score, accuracy, and ROC values. We evaluate the efficiency of the MFO-RELM method to other machine learning methods such as naive Bayes, logistic regression, support vector machines, and k-nearest neighbor.