Cliff Walking with Q-Learning NetLogo Extension (version 1.0.0)
This model implements a classic scenario used in Reinforcement Learning problem, the “Cliff Walking Problem”. Consider the gridworld shown below. This is a standard undiscounted, episodic task, with start and goal states, and the usual actions causing movement up, down, right, and left. Reward is -1 on all transitions except those into the region marked “The Cliff.” Stepping into this region incurs a reward of -100 and sends the agent instantly back to the start (SUTTON; BARTO, 2018).
The problem is solved in this model using the Q-Learning algorithm. The algorithm is implemented with the support of the NetLogo Q-Learning Extension
Release Notes
This model implements Q-learning (Watkins 1989) a one-step temporal difference algorithm in the area of reinforcement learning, a branch of artificial intelligence and machine learning.
The scenario used is a classic Reinforcement Learning problem, the “Cliff Walking”. Consider the gridworld shown below. This is a standard undiscounted, episodic task, with start and goal states, and the usual actions causing movement up, down, right, and left. Reward is -1 on all transitions except those into the region marked “The Cliff.” Stepping into this region incurs a reward of -100 and sends the agent instantly back to the start (SUTTON; BARTO, 2018).
This implementation makes use of the Q-Learning NetLogo Extension to implement the Q-Learning algorithm.
This is a companion discussion topic for the original entry at https://www.comses.net/codebases/b938a820-f209-4648-afc6-0946657c3484/releases/1.0.0/