Inverse Noncooperative Dynamic Games
Timothy L. Molloy, Jason Ford, Tristán Pérez
Abstract
Timothy L. Molloy, Jason Ford, Tristán Pérez
Abstract
We consider the problem of computing parameters of player cost functions in discrete-time nonzero-sum noncooperative dynamic games from open-loop Nash equilibria. Although similar inverse problems have been investigated in the optimal control literature where there is a single player (or decision maker), there has been limited attention given to the inverse dynamic game problem with multiple (competing) players. By exploiting the minimum principle of optimal control, we propose a method of inverse dynamic games for when the information structure of the game is open-loop. Our method involves solving a system of linear equations and is able to recover the true unknown parameters (up to an unknown scaling factor) whenever a testable rank condition holds. We illustrate our method in an example two-player game.
OpenAlex reports 17 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We consider the problem of computing parameters of player cost functions in discrete-time nonzero-sum noncooperative dynamic games from open-loop Nash equilibria. Although similar inverse problems have been investigated in the optimal control literature where there is a single player (or decision maker), there has been limited attention given to the inverse dynamic game problem with multiple (competing) players. By exploiting the minimum principle of optimal control, we propose a method of inverse dynamic games for when the information structure of the game is open-loop. Our method involves solving a system of linear equations and is able to recover the true unknown parameters (up to an unknown scaling factor) whenever a testable rank condition holds. We illustrate our method in an example two-player game.
Key concepts: Sequential game, Inverse, Mathematical optimization, Rank (graph theory), Computer science, Nash equilibrium, Mathematical economics, Repeated game