Reinforcement Learning Hierarchical Neuro-Fuzzy Politree Model for Control of Autonomous Agents

This work presents a new hybrid neuro-fuzzy model for automatic learning of actions taken by agents. The main objective of this new model is to provide an agent with intelligence, making it capable, by interacting with its environment, to acquire and retain knowledge for reasoning (infer an action). This new model, named Reinforcement Learning Hierarchical Neuro-Fuzzy Politree (RL-HNFP), and its improved version (RL-HNFP + ), descends from the Reinforcement Learning Hierarchical Neuro-Fuzzy BSP (RL-HNFB) that applies binary hierarchical partitioning methods, together with the Reinforcement Learning (RL) methodology. These two characteristics permit the autonomous agent to automatically learn its structure and its necessary action in each position in the environment. The RL-HNFP model and its extension (RL-HNFP + ) are evaluated in a well known benchmark application in the area of autonomous agents: the Mountain Car Problem. The results obtained demonstrate the potential of these models, which operate without any prior information, such as number of rules, rules specification, or number of partitions that the input space should have, providing good performance and demonstrating the agent’s autonomy.