Autoregressive Modeling and Prediction of Annual Worldwide Cybercrimes for Cloud Environments

Abstract

Recently, cybercrimes are causing huge impact on different cyber systems that might include vital information such as financial transactions and medical records. A better understanding of the accelerating numbers of cybercrimes and their enormous cost could help the global in bridging the gap between their defenses and the escalating numbers cyber criminals. In this paper, we present an estimation model of cybercrimes time series using auto-regressive (AR) model by employing the optimal modeling order that maximizes the estimation accuracy while maintaining minimum prediction error. The proposed model was developed using Matlab to estimate the time series for yearly global number of cybersecurity incidents activity during the period from 2009-2018 and forecast the figures for next upcoming years 2019-2020. The simulation results showed that the optimal model order to estimate the given cybercrime activity is AR(4) since its corresponds to minimum acceptable predication error values to estimate the signal recording an estimation accuracy of 93.5%.