Automated Detection and Analysis using Mathematical Calculations

A compromised system usually shows some form of anomalous behaviour. Examples include new processes, services, or outbound traffic. In an ideal environment, rules are configured to alert on such anomalies, where an analyst would perform further analysis to determine a possible compromise. However, the real-world situation is less than ideal; new processes, outbound traffic, or other anomalies often blend into legitimate activities. A large network can generate terabytes of data daily, causing the task of developing efficient detection capabilities a bit challenging. Mathematical calculations can enhance detection capability by emulating the human confidence level on assessment and analysis. Mathematical analysis can help understand the context of the event, establishing fidelity of the initial investigation automatically. By incorporating automated analysis to handle false positives, human errors and false negative can be avoided, resulting in a greater detection and monitoring capability.