DRU: Network Structure, Behavioral Considerations and Risk Management in Interdependent Security Games
PI/Project Contact: Michael Kearns and
We propose a broad program of research to significantly grow the theory and applicability of interdependent security (IDS) games, a recently introduced class of models that captures a wide variety of collective risk and decision-making scenarios that include problems in airline security, corporate governance, computer network security and vaccinations against diseases. To expand our knowledge and experience with IDS models we propose research projects in three areas:
- Effects of Network Structure on IDS Outcomes, Computation and Learning. Here we propose a mixture of mathematical and experimental work. This includes an investigation of the marriage of IDS models with network formation models from the social network theory literature, and a large-scale numerical application of IDS models to problems in network security (on which we will be joined by Prof. Matt Blaze of Penn). We also propose developing computational methods for manipulating IDS models and developing more efficient solution procedures for large-scale problems.
- Behavioral IDS Models and Experiments. Here we propose a program of behavioral and experimental work to examine how human participants actually behave in IDS settings. We propose to develop descriptive models of choice by individuals in an IDS environment and will specify hypotheses to be tested through controlled experiments.
- Risk Management Strategies and Policy Implications To encourage and induce individuals to invest in cost-effective protection measures for IDS problems, we will examine several risk management strategies that include providing risk information, communication with others, economic incentives, and tipping strategies. These strategies are designed to foster cooperative behavior between individuals.
The proposed research examines a class of models for important risk and decision-making problems that are commonly faced by individuals, organizations, and nations. The research focuses on fundamental issues associated with network structure and behavioral considerations, both of which impact IDS modeling. The proposed advances in computational methods will permit the application of IDS models to large-scale problems. By examining alternative risk management strategies that are guided by how individuals make protective decisions in an IDS environment, we will link prescriptions with descriptive models of choice.
The proposed work is interdisciplinary in nature and should serve as an exciting focal point for researchers in computer science, decision and management sciences, economics, psychology, risk management and policy analysis. Interdependent security (IDS) models require one to address the likelihood of certain events occurring and the ways in which their consequences are propagated throughout entire systems. One needs to focus on weakest links in interdependent supply chains and systems that have broad implications for developing risk management strategies for important societal problems.