.. image:: pystat.png
    :height: 20
    :alt: Statistique
    :target: http://www.xavierdupre.fr/app/ensae_teaching_cs/helpsphinx/td_2a_notions.html#pour-un-profil-plutot-data-scientist

.. _l-td2a-reinforcement-learning-marl:

Multi Agent Reinforcement Learning (MARL)
+++++++++++++++++++++++++++++++++++++++++

*Lectures*

* `Decentralised Learning in Systems with Many, Many Strategic Agents <https://arxiv.org/pdf/1803.05028.pdf>`_
* `Developing a Python Reinforcement Learning Library for Traffic Simulation <http://ala2017.it.nuigalway.ie/papers/ALA2017_Ramos.pdf>`_
* `Multiagent Reinforcement Learning: Theoretical Framework and an Algorithm <http://www.lirmm.fr/~jq/Cours/3cycle/module/HuWellman98icml.pdf>`_
* `Reinforcement Learning and Nonparametric Detection of Game-Theoretic Equilibrium Play in Social Networks <https://arxiv.org/pdf/1501.01209.pdf>`_
* `Efficient Learning Equilibrium <https://ie.technion.ac.il/~moshet/ele-journal-revised3.pdf>`_
* `Paper Collection of Multi-Agent Reinforcement Learning (MARL) <https://github.com/LantaoYu/MARL-Papers>`_
* `A reinforcement learning algorithm for building collaboration in multi-agent systems <https://arxiv.org/pdf/1711.10574.pdf>`_

*Modules*

* `pyrl <https://github.com/goramos/pyrl>`_ (pas vraiment finie)
* `streamingbandit <https://github.com/Nth-iteration-labs/streamingbandit>`_

*Environnements*

* `OpenAI Gym <https://gym.openai.com/>`_, l'outil propose une
  formalisation qui permet de tester
  les algorithmes d'apprentissage par renforcements pour ses propres
  expériences ou pour des contextes ou jeu prédéfinies.
  Cela peut aboutir à ce type d'expérience :
  `OpenAI Gym for NES games + DQN with Keras to learn Mario Bros. from raw pixels <https://naereen.github.io/gym-nes-mario-bros/>`_.