Are You Doing What I Think You Are Doing? Criticising Uncertain Agent Models
This addresses the lack of a universal theory for critiquing uncertain agent models in multiagent applications, though it appears incremental as it builds on existing methods for hypothesis construction.
The paper tackles the problem of evaluating the correctness of behavioral hypotheses in multiagent systems by introducing a novel frequentist hypothesis test algorithm that learns the test statistic distribution during interaction with asymptotic correctness guarantees. The results show the algorithm achieves high accuracy and scalability at low computational costs.
The key for effective interaction in many multiagent applications is to reason explicitly about the behaviour of other agents, in the form of a hypothesised behaviour. While there exist several methods for the construction of a behavioural hypothesis, there is currently no universal theory which would allow an agent to contemplate the correctness of a hypothesis. In this work, we present a novel algorithm which decides this question in the form of a frequentist hypothesis test. The algorithm allows for multiple metrics in the construction of the test statistic and learns its distribution during the interaction process, with asymptotic correctness guarantees. We present results from a comprehensive set of experiments, demonstrating that the algorithm achieves high accuracy and scalability at low computational costs.