PREVE: A Policy Recommendation Engine based on Vector Equilibria Applied to Reducing LeT's Attacks

摘要

We consider the problem of dealing with the terrorist group Lashkar-e-Taiba (LeT), responsible for the 2008 Mumbai attacks, as a five-player game. However, as different experts vary in their assessment of players' payoffs in this game (and other games), we identify multi-payoff equilibria through a novel combination of vector payoffs and well-supported ε approximate equilibria. We develop a grid search algorithm for computing such equilibria, and provide experimental validation using three payoff matrices filled in by experts in India-Pakistan relations. The resulting system, called PREVE, allows us to analyze the equilibria thus generated and suggest policies to reduce attacks by LeT. We briefly discuss the suggested policies and identify their strengths and weaknesses.