A Complexity Analysis of Cooperative Mechanisms in Reinforcement Learning

By Steven D. Whitehead. 1991

How researchers see this paper

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

Citations in the corpus

Is Independent Learning All You Need in the StarCraft Multi-Agent Challenge? (2020, background): “ction approximators were subsequently developed [IAC, deep IQL][19, 9]. The question of whether independent agents are better at learning cooperative behaviour than multiple complete observing agents [33] was ﬁrst comprehensively investigated by Tan [27], who concludes that while sharing policies or”
Multiagent Learning: Basics, Challenges, and Prospects (2012, background): “Therefore extensions of the MDP framework have been considered such as Markov games and joint action learners (Littman 1994, Claus and Boutilier 1998).”
Half Field Offense in RoboCup Soccer: A Multiagent Reinforcement Learning Case Study (2006, methodology): “Whitehead [13] describes a Learn-by-Watching method similar to ours and obtains theoretical bounds for the speedup in learning by multiple Q-learning agents.”
AUTOMATED DISCOVERY OF OPTIONS IN REINFORCEMENT LEARNING (2003, background): “In such cases, an exponential number of time steps before reaching the goal is to be expected as is documented in the literature [34][35].”
Multi Agent Reinforcement Learning Independent vs Cooperative Agents (2003, background): “This result demonstrates another beneet of learning in a cooperative society w h e r e n o vices can learn quickly from experts by examples (Lin 1991, Whitehead 1991).”
Accelerating Reinforcement Learning through Implicit Imitation (2003, methodology): “Our model is also distinct from models of explicit teaching (Lin, 1992; Whitehead, 1991b): we do not assume that the mentor has any incentive to move through its environment in a way that explicitly guides the learner to explore its own environment and action space more effectively.”
Purposive behavior acquisition for a real robot by vision-based reinforcement learning (1995, methodology): “In the Learning with an External Critic (or LEC) [ 11 ], the robot receives an advise in each state from the external critic.”
Parallel reinforcement learning with linear function approximation (2007, unknown): “While there have been preliminary analyses [3, 7] demonstrating the promise of a parallel approach to RL, the cost of inter-agent communication has usually been ignored.”
Using communication to reduce locality in distributed multiagent learning (1997, methodology): “Whitehead (1991) analyzed cooperative RL mechanismsand Littman (1994) used the Markov games framework for simulated RLsoccer-playing agents.”
Autonomous Agents and Multi-agent Systems Manuscript No. Learning Potential Functions and Their Representations for Multi-task Reinforcement Learning (unknown): “In RL, it may refer either to training the agent on successive tasks of increasing complexity, until the desired complexity is reached [17,27,57,59,61,65], or, more commonly, to supplementing the MDP’s reward function with additional, artificial rewards [3,14,15,24,38,49,50,53,85].”
Reinforcement Learning: an Overview (2000, background): “The difficulty is double: – the agent must not lose of view the final objective(2); – the agent can be confused when this intermediate signal disappears, problem identified by Whitehead [34] and known as extinction effect.”
Off-Policy Policy Search (2007, background): “Firs t Whitehead showed that the expected length of theQL’s first trial can grow exponentially with the size of the envi ronment [10].”