Proposition of multi-agent conflict situation simulation and reinforcement learning toolset with demonstration of early stage implementation
Robert Jarosz · Computer Science and Mathematical Modelling · 2024
This paper introduces conceptual approach to modelling conflicts. A flexible framework compatible in development phase is presented. Model scalability, possibility of parallelization and computational distribution over network is discussed. As example of application there are presented two variants of classic game theory problems. At the end of the paper current problems are briefly stated and future work direction is presented.