The world of independent learners is not markovian

By Guillaume J. Laurent, Laëtitia Matignon, Nadine Le Fort-Piat. 2011

123 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Management Science and Operations Research, Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Management Science and Operations Research

In Statistical Learning Theory

Citations in the corpus

Value-Decomposition Networks For Cooperative Multi-Agent Learning (2017, background): “In general each agent is then faced with a non-stationary learning problem because the dynamics of its environment effectively changes as teammates change their behaviours through learning (Laurent et al., 2011).”
Deep Decentralized Multi-task Multi-Agent Reinforcement Learning under Partial Observability (2017, background): “However, the domain appears non-stationary from the perspective of each Dec-POMDP agent, a property we formalize by extending the deﬁnition by Laurent et al. (2011 (2) Letting π − ( i ) = π \ { π ( i ) } , non-stationarity from the lo-cal perspective of agent i follows as in general a , which…”
Independent reinforcement learners in cooperative Markov games: a survey regarding coordination problems (2012, background): “Moreover, it can cause the ‘destruction’ of the optimal policy of an IL (Laurent et al., 2010).”
A Unified Game-Theoretic Approach to Multiagent Reinforcement Learning (2017, background): “Aside from the problem that these local environments are non-stationary and non-Markovian [57] resulting in a loss of convergence guarantees for many algorithms, the policies found can overfit to the other agents’ policies and hence not generalize well.”
Value-Decomposition Networks For Cooperative Multi-Agent Learning Based On Team Reward (2018, background): “In general, each agent is then faced with a non-stationary learning problem because the dynamics of its environment effectively changes as teammates change their behaviours through learning [14].”
Emergence of Scenario-Appropriate Collaborative Behaviors for Teams of Robotic Bodyguards (2018, background): “InRL agents treat the interactions of other agents as a part of their local environment, which makes them non-stationary and non-Markovian (Laurent et al. 2011).”
Multi-agent deep reinforcement learning: a survey (2021, background): “As a result, the Markov assumption is violated, and the learning constitutes an inherently difficult problem (Hernandez-Leal et al. 2017; Laurent et al. 2011).”
A multi-agent reinforcement learning model of common-pool resource appropriation (2017, methodology): “These algorithms were developed for the single agent case and are applied independently here [21, 3] even though this multi-agent context breaks the Markov assumption [20].”