Multi-Agent Exploration for Faster and Reliable Deep Q-Learning Convergence in Reinforcement Learning
Abhijit Majumdar, Patrick J. Benavidez, Mo M. Jamshidi · 2018
Function approximation based Q-learning, using deep q-learning has had recent extraordinary developments applicable to generalized applications. Many techniques have been introduced to counter the inherent caveats in using a deep neural network in reinforcement learning. We demonstrate the use of multi-agent virtual exploration integrated into existing algorithms to show better convergence property, and show how they can be applied as extensions to provide faster and better converged values.