Cyber-Crime Detection: Experimental Techniques Comparison Analysis

Abstract

Cyber-crime is one of the main problems the world face, and machine learning plays a key part in contemporary operating systems for giving better transformation in the security environment and cybercrime detection. While detecting cybercrimes is difficult, it is possible to gain advantages from machine learning to generate models to assist in predicting and detecting cybercrimes. The researchers have proven that the majority of the models can work effectively in identifying cybercrime, they can span from 70% to 90% in accuracy measuring. The objective of this research paper is to conduct experimental techniques comparison analysis for cyber-crime detection by reviewing all possible machine learning algorithms for automatic detection. The key focus of the study is on the use of eight classifiers models which are Logistic Regression (LR), Decision Tree (DT), K-nearest Neighbors (KNN), Support Vector Machine (SVM), Naive Bayes (NB), Random Forest (RF), eXtreme Gradient Boosting (XGBoost) and Multiple layer perception (MLP). From the experiment conducted, the high prediction came from MLP which is 96% accuracy of the cyber-crime methods based on existing cyber-crime data.