Fast Reinforcement Learning Techniques Using the Euclidean Distance and Agent State Occurrence Frequency.
Hani Al-Dayaa, Dalila B. Megherbi · 2006
Abstract – Reinforcement learning techniques like the Q-Learning one and the Multiple-Lookahead-Levels one that we introduced in our prior work require the agent to complete an initial exploratory path followed by as many hypothetical and physical paths as necessary to find the optimal path to the goal. This paper introduces a reinforcement learning technique that uses the Euclidean distance to the goal as a main measure for an autonomous agent’s action selection. As we show in this paper, no exploratory or hypothetical paths are required, there is no need to cover all possible states, and the agent requires a maximum of two physical paths to find the optimal path to the goal. The agent’s state occurrence frequency is introduced here and used to support the proposed Distance-Only technique. A computation speed performance analysis is carried out, and the Distance-and Frequency technique is shown to require less computation time than the Q-Learning one.