Enhancing Asset Security in Malaysia: A Multivariate Regression and Time Series Analysis Approach

Abstract

This research addresses the challenge of managing and securing external assets in an organization’s digital infrastructure. As the attack surface grows due to factors like software updates, configuration changes, outdated security policies, and the addition of new assets, organizations become increasingly vulnerable to security threats. Without proper forecasting and analysis, these trends can lead to inefficiencies in resource allocation and expose critical assets to cyberattacks. the Autoregressive Integrated Moving Average model was employed to forecast changes in the external attack surface and prediction the quantity of total assets exposed over time. Next, multivariate linear regression was used to analyse the relationships between various factors. Influence diagrams was used to visualize the different factors, uncertainties, and decisions interact in the context of resource allocation and security planning. The results presented those certain factors, such as frequent software updates and the addition of new assets, significantly contributed to the expansion of the attack surface. Then, the strongest predictors of asset exposure were identified, which allowed for more targeted interventions. The influence diagrams provided a clear, visual representation of how these factors interact, aiding in the understanding of complex security scenarios. This research  analysing critical relationships with multivariate linear regression, organizations can better allocate resources to mitigate risks.