My work in the field of Artificial Intelligence (AI) is motivated by real-world societal problems, particularly in the areas of Public Safety and Security (e.g., urban crime prevention and counterterrorism), Cybersecurity (e.g., the protection of network data from stealthy botnets), Sustainability (e.g., wildlife and fish protection), and Public Health (e.g., the prevention of vaccination misinformation spread on social networks). I aim to bridge the gap between theory and application in AI by providing practical, computational, AI-based solutions for these problems. The solutions I have developed employ techniques drawn not only from AI's various subfields, including Multi-Agent Systems, Game Theory, Machine Learning, and Optimization, but also from fields outside of AI, such as Cognitive Modeling and Conservation Biology.
My work is being successfully utilized in the real world: my models and algorithms have been incorporated into the wildlife-protection application PAWS (Protection Assistant for Wildlife Security), which has been extensively used by NGOs such as Panthera and the Wildlife Conversation Society in conservation areas in Malaysia and Uganda.
In many problems in domains of security and sustainability, human attackers are boundedly rational, and often attack non-optimal targets. As an example, in wildlife protection, rangers often find poaching signs spread over multiple locations (with different animal density) in the conservation area.
This project focuses on developing new behavioral models to predict the decision making of the attackers in these adversarial domains based on (i) human subject data; and (ii) real-world attack data. We employ techniques drawn from various fields, including Machine Learning, Game Theory, and Cognitive Modeling.
Real-world security domains are often characterized by partial information: uncertainty (particularly on the defender’s part) about actions taken or underlying characteristics of the opposing agent. Experience observed through repeated interaction in such domains provides an opportunity for the defender to learn about the behaviors and characteristics of attacker(s).
To the extent that the defender relies on data, however, the attacker may choose to modify its behavior to mislead the defender. That is, in a particular interaction the attacker may select an action that does not actually yield the best immediate reward, to avoid revealing sensitive private information. Such deceptive behavior could manipulate the outcome of learning to the long-term benefit of the attacker. This project investigates strategic deception on the part of an attacker with private information.
This project focuses on the development of practical game-theoretic solutions for real-world complex large-scale cybersecurity domains. Cybersecurity problems often involve dynamic stochastic interactions between network security administrators and cybercriminals on a computer network with exponential action spaces, imperfect information of players' knowledge and actions, and several potential unforeseen uncertainties. We are interested in applying simulation-based methodologies, particularly empirical game-theoretic analysis, and exploring a variety of parameterized heuristic game-theoretic solutions. This approach allows me to model and analyze complex cybersecurity scenarios.
Negotiation is an important skill for any social entity (human or machine). Of course, negotiation is ubiquitous in economic transactions, but negotiation happens across a wide array of contexts, even when people don't recognize that they are negotiating. People negotiate over what movie to see. To obtain a job, people must negotiate a salary and job responsibilities. This project studies deceptive strategies of negotiators, specifically, how they might best accomplish this aim.
Many adversarial domains such as counterinsurgency exhibits contagious actions of competitors. In public health, in particular, there exist anti-vaccination groups who are intentionally spreading vaccination misinformation on social networks and thus health experts need to figure out an efficient way to fight against anti-vaccination messages. This project studies game-theoretic models and algorithms to control contagion on a social network of which outcomes are stochastic.