Recognizing the enemy: Combining reinforcement learning with case based reasoning in domination games

This thesis evaluates the benefits of using Cased Based Reasoning with Reinforcement Learning to identify and learn strategies in Domination Games. It will show that Case Based Reasoning can speed up the Reinforcement Learning process when facing opponents similar to those seen in the past. Along with this the paper will also show that reinforcement learning can be applied effectively in multiple Domination Game domains through introducing the new Domination Game Simulator DOM.