The hierarchical interaction between the actor and critic in actor-critic based reinforcement learning algorithms naturally lends itself to a game-theoretic interpretation. We adopt this viewpoint model as two-player general-sum game with leader-follower structure known Stackelberg game. Given abstraction, we propose meta-framework for where leader player follows total derivative of its objecti...

Simultaneous (Nash) and sequential (Stackelberg) equilibria of two-player dynamic quadratic cheap talk and signaling game problems are investigated under a perfect Bayesian formulation. For the dynamic scalar cheap talk, a zero-delay communication setup is considered for i.i.d. and Markov sources; it is shown that the final stage equilibrium is always quantized and under further restrictive con...

In a recent paper Bagwell (1995) pointed out that only the Cournot outcome, but not the Stackelberg outcome, can be supported by a pure Nash equilibrium when actions of the Stackelberg leader are observed with the slightest error. The Stackelberg outcome, however, remains close to the outcome of a mixed equilibrium. We compare the predictions in various classes of evolutionary and learning proc...

In order to address inefficiencies of Nash equilibria for horizontal queuing congestion networks, we study the Stackelberg routing game on parallel networks: assuming a coordinator has control over a fraction of the flow, what is a good way of routing that compliant flow so that the induced Nash equilibrium is closer to the social optimum than the initial Nash equilibrium? We study Stackelberg ...

The continuous network design problem (CNDP) is known to be difficult to solve due to the property of non-convexity and nonlinearity. Such kind of CNDP can be formulated as a bi-level programme, which may be classified into Stackelberg approach and Nash one according to the relationship between the upper level and the lower level parts. This paper formulates the CNDP for road expansion based on...

In a recent paper Bagwell (1995) pointed out that only the Cournot outcome, but not the Stackelberg outcome, can be supported by a pure Nash equilibrium when actions of the Stackelberg leader are observed with the slightest error. The Stackelberg outcome, however, remains close to the outcome of a mixed equilibrium. We compare the predictions in various classes of evolutionary and learning proc...

• We introduce a multistage Stackelberg game for inspections with multiple targets. propose two solutions methods, using backward induction and linear programming. provide an explicit solution inspection where one inspector has the role of monitoring group inspectees. The resources to visit only few them. Visits are performed sequentially no repetitions. inspectees report share sequence as they...