Skip to content

Introduce reinforcement learning agent #7

Description

@Johansmm

Based of reinforcement learning theory, introduce an agent that perform an action only with the current state of the game.
For that, it should be trained before to play, using the following knowledged:

The goal will be to find an RL-based agent that competes against greedy algorithm, and that reach to win more than 80% of the time. Different learning functions (Q-learning, policy, mix, others) can be tried.

Note: a brief summary could be regard on: https://arxiv.org/pdf/1708.05866.pdf

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions