An Anticipatory Rapid-Response Digital Threat Evaluation Framework Combining Stochastic State Modeling With Probabilistic Reasoning Structures

This paper presents a proactive and time-sensitive cyber risk assessment model that integrates Markov chains and Bayesian networks to evaluate cybersecurity risks dynamically. The system leverages the Exploit Prediction Scoring System (EPSS) to estimate short-term exploitation probabilities and uses Bayesian networks to model asset interdependencies. Attack paths are identified using Depth-First Search (DFS) and evaluated using absorbing Markov chains to compute probabilistic risk distributions. The model provides time-to-compromise insights and supports proactive decision-making through dynamic risk visualization. The approach enhances traditional static methods by offering predictive, time-aware, and actionable cybersecurity analysis.