Modeling Parameters and Techniques for Risk Modeling

Abstract

To make more informed cyber-assurance decisions, the information security community needs better models to evaluate and predict future adversarial behavior and greatest return-on-investment in defensive measures. In addition, adversary characterization information on their success rate, resources expended, risk perception, and attack decision trade-off preferences are almost always derived from observable characteristics and rarely directly acquired. Rather than abdicate use of analytical techniques or resort to using less rigorous modeling techniques because of a lack of available data, this chapter proposes a set of techniques to be used to capture the best available information at the time for use in more rigorous models, and provides insight into the level of confidence the authors have in this data. From this information, one can also then derive modeling estimation parameters to meet the need for specific confidence levels. It can also provide us guidance on how to invest to improve our data estimates.