Iterative Model-Based Reinforcement Learning Using Simulations in the Differentiable Neural Computer
This work addresses the challenge of continual learning in reinforcement learning for agents operating in dynamic environments, though it appears incremental as it builds on existing DNC and PPO methods.
The authors tackled the problem of lifelong reinforcement learning by pairing an agent with a Differentiable Neural Computer (DNC) model to iteratively train in simulations, achieving successful task-solving in synthetic environments using Proximal Policy Optimization.
We propose a lifelong learning architecture, the Neural Computer Agent (NCA), where a Reinforcement Learning agent is paired with a predictive model of the environment learned by a Differentiable Neural Computer (DNC). The agent and DNC model are trained in conjunction iteratively. The agent improves its policy in simulations generated by the DNC model and rolls out the policy to the live environment, collecting experiences in new portions or tasks of the environment for further learning. Experiments in two synthetic environments show that DNC models can continually learn from pixels alone to simulate new tasks as they are encountered by the agent, while the agents can be successfully trained to solve the tasks using Proximal Policy Optimization entirely in simulations.