Context-specific multiagent coordination and planning with factored MDPs

By Carlos Guestrin, Shobha Venkataraman, Daphne Koller. 2002

138 citations recorded by the source corpus.

Subfields: Reinforcement Learning, Game & Decision Theory

How researchers see this paper

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

In Reinforcement Learning

In Game & Decision Theory

Citations in the corpus

Deep Coordination Graphs (2019, methodology): “An undirected coordination graph (CG, Guestrin et al., 2002a) G = (cid:104)V , E(cid:105) contains a vertex v i ∈ V for each agent 1 ≤ i ≤ n and a set of undirected edges { i, j } ∈ E between vertices v i and v j .”
Value Function Transfer for Deep Multi-Agent Reinforcement Learning Based on N-Step Returns (2019, background): “Earlier work focuses on constructing coordination graphs (CGs) to represent the interactive relationship between agents and conducting learning according to the speciﬁed relationship in each state [Guestrin et al. , 2002a; Guestrin et al. , 2002b; Kok and Vlassis, 2004 ] .”
Multi-Agent Common Knowledge Reinforcement Learning (2018, background): “Guestrin et al. (2002b) represent agents’ value functions as a sum of context-speciﬁc value rules that are part of the agents’ ﬁxed a priori common knowledge.”
Efficient Solution Algorithms for Factored MDPs (2003, positive): “This paper is a greatly expanded version of work that was published before in Guestrin et al. (2001a), and some of the work presented in Guestrin et al. (2001b, 2002) .”
Decentralized MDPs with sparse interactions (2011, background): “Coordination graphs [6,43] capture local dependences between the different agents in an MMDP, allowing the overall Q function to be decomposed into local Q -functions that can be optimized individually.”
Learning in Multi-agent Systems with Sparse Interactions by Knowledge Transfer and Game Abstraction (2015, methodology): “Earlier work makes use of coordination graphs to specify the coordination dependencies between agents in particular states [7, 8, 14].”
Learning of coordination: exploiting sparse interactions in multiagent systems (2009, background): “In [8, 9], the authors describe local interactions between the agents using coordination graphs.”
Solving Transition Independent Decentralized Markov Decision Processes (2004, background): “Guestrin, Venkataraman, and Koller (2002) focused on interactions between agents through the reward function, similar to what we do here, however their observation model more resembles the full observability of the MMDP than the partial observability of our model.”
Sparse cooperative Q-learning (2004, methodology): “At the other extreme, we have the independent learners (IL) approach (Claus & Boutilier, 1998) in which the agents ignore the actions and rewards of the other agents in the system, and learn their strategies independently.”
Utile Coordination: Learning Interdependencies Among Cooperative Agents (2005, methodology): “We refer to [Guestrin et al., 2002b] for details.”
Generalizing plans to new environments in relational MDPs (2003, background): “[7] show thatcontext-specific independence an allow for dynamically changing coordination structures in multiagent environme ts.”
Graphical Model Inference in Optimal Control of Stochastic Multi-Agent Systems (2008, background): “Guestrin, Koller, and Parr (2002a) and Guestrin, Venkataraman, and Koller (2002b) assumed a predefined approximate structure of the value functions, and thereby provided an efficient approximate MDP model for multi-agent systems.”