Solving bilateral bargaining with arbitrary one-sided uncertainty
DEI - 3B Room
November 18th, 2010
Bargaining is one of the most important negotiation problems and its resolution is an important step towards the automation of electronic transactions. According to the bilateral setting, bargaining is modeled as a non-cooperative extensive-form game with infinite actions (offers are real values) where two agents with conflicting objectives must reach an agreement. The behaviour of each agent is modeled by a set of parameters. Notwithstanding its prominence, literature does not present a satisfactory solution in the presence of uncertainty over multiple parameters. We propose an algorithm that reduces a bargaining problem to a finite game, solves this last game, and then maps its strategies to the original continuous game. Computational results show that the required time increases exponentially with the dimension of the game and only small cases can be solved exactly. This motivates the need to resort in future works to concepts of approximate equilibrium and to abstractions for reducing the size of the game tree.
Applied systems analysis and operations research