Decision making under uncertainty is a feature that is present in all aspects of homeland security applications. The “Unity of Effort memo” introduced by Secretary Johnson calls for a sound decision making structure across all components of DHS.
The Decision Analysis research area of CREATE is concerned with the use of rigorous models to help make sound decisions in homeland security applications based on axioms of rational thought to help provide a “unity of effort” decision culture.
Topics in this area include:
- The appropriate representation of preferences and trade-offs using multiattribute value and utility models.
- The appropriate representation of risk attitudes.
- Modeling adversary beliefs and preferences
- The representation of belief using joint probability distributions, and the updating of belief using Bayes’ rule.
- Group decision making
- Framing decision situations
- Alternative generation
- The identification (and minimization) of the effects of cognitive biases.
- Implementing a decision culture within an enterprise.
The Decision Analysis area also include aspects of learning from data, and the steps involved in the transfer of data to knowledge and ultimately to decision making.
Security is a critical concern around the world, whether it is the challenge of protecting ports, airports and other critical national infrastructure, or protecting wildlife/forests and fisheries, or suppressing crime in urban areas. In many of these cases, limited security resources prevent full security coverage at all times. Instead, these limited resources must be allocated and scheduled efficiently, avoiding predictability, while simultaneously taking into account an adversary's response to the security coverage, the adversary's preferences and potential uncertainty over such preferences and capabilities.
Computational game theory can help us build decision-aids for such efficient security resource allocation. Indeed, casting the security allocation problem as a Bayesian Stackelberg game, we have developed new algorithms that are deployed over multiple years in multiple applications including:
Fundamentally, we are focused on the research challenges in these efforts, marrying these applications with research on topics such as (i) fast algorithms for solving massive-scale games; (ii) behavioral game theory research for addressing human adversaries who may act with bounded rationality and imperfect observations; (iii) understanding the impact of players' limited observations on solution approaches adopted.
Risk Management plays a central role in DHS strategic planning. DHS has a long tradition of risk management but requires new analytic methods to address contemporary homeland security issues across multiple threat domains. The application of analytic methods to improve decision-making is the domain of Operations Research. CREATE combines Risk Management and Operations Research to develop new methods to address three key analytical challenges. First, many problems that DHS programs are intended to solve involve adversarial competitions. This means that DHS must consider the benefits of its programs against both current risks and anticipated shifts in risks as new threats emerge of different types and in different places. Second, because this dynamic threat environment creates many types of risk in many places, it is not appropriate to evaluate security programs in isolation. Instead, DHS programs must be evaluated within the context of how they contribute to a risk management system designed to achieve multiple goals. Third, though risk management systems are designed to counter many types of threats, it is challenging for DHS to prevent adversaries from succeeding in all scenarios. Thus, an important aspect of risk management across DHS is the systematic incorporation of means of deterrence and dissuasion to influence adversary decisions so terrorists act in ways that are easier to detect, more difficult for them to succeed, or have easier consequence to accept.
The primary goal of this theme is to use Operations Research approaches to improve the understanding of valuation of security benefits across threat domains, the development, evaluation and integration of risk management and resource allocation applications in collaboration with DHS partners, and ultimately, to enable better, faster, more focused policy- and decision-making.