Predicting Cyber Threats Using Machine Learning for Improving Cyber Supply Chain Security
Areej Omar Al-Ansari, Tahani Alsubait · 2022
Recent history has shown many cases where organizations have been attacked through less secure points in their supply chains. Supply chains are defined as a series of steps taken to obtain a product or service for a customer. Cybersecurity threats target the most vulnerable points in these organizations to inflict damage to them. The cascading effect that can result from the complexity of supply chains has made them an ambitious target for hackers to reach more victims and information, whether from people or organizations. Therefore, there was a need to predict threats that can affect supply chains using machine learning and artificial intelligence techniques to reduce these threats and put in place the best controls as a proactive step that would reduce and avoid the consequences. In this research, Microsoft Malware Predictions dataset was used to predict cyber threats for cyber supply chains. Five machine learning algorithms were used to predict threats, including knearest neighbor algorithm, support vector machine, logistic regression, random forest, and LightGBM. The results show that out of the five algorithms used, Random Forest algorithm and LightGBM algorithm have good and high accuracy with 72% compared to the other three, followed by Logistic Regression algorithm and Support Vector Machine algorithm and finally kNearest Neighbor algorithm which is less accurate than the other algorithms.