๐Ÿ˜‡
Deep Multi-Agent Reinforcement Learning
  • Deep Multi-Agent Reinforcement Learning
  • Abstract & Contents
    • Abstract
  • 1. Introduction
    • 1. INTRODUCTION
      • 1.1 The Industrial Revolution, Cognition, and Computers
      • 1.2 Deep Multi-Agent Reinforcement-Learning
      • 1.3 Overall Structure
  • 2. Background
    • 2. BACKGROUND
      • 2.1 Reinforcement Learning
      • 2.2 Multi-Agent Settings
      • 2.3 Centralized vs Decentralized Control
      • 2.4 Cooperative, Zero-sum, and General-Sum
      • 2.5 Partial Observability
      • 2.6 Centralized Training, Decentralized Execution
      • 2.7 Value Functions
      • 2.8 Nash Equilibria
      • 2.9 Deep Learning for MARL
      • 2.10 Q-Learning and DQN
      • 2.11 Reinforce and Actor-Critic
  • I Learning to Collaborate
    • 3. Counterfactual Multi-Agent Policy Gradients
      • 3.1 Introduction
      • 3.2 Related Work
      • 3.3 Multi-Agent StarCraft Micromanagement
      • 3.4 Methods
        • 3.4.1 Independent Actor-Critic
        • 3.4.2 Counterfactual Multi-Agent Policy Gradients
        • 3.4.2.1 baseline lemma
        • 3.4.2.2 COMA Algorithm
      • 3.5 Results
      • 3.6 Conclusions & Future Work
    • 4 Multi-Agent Common Knowledge Reinforcement Learning
      • 4.1 Introduction
      • 4.2 Related Work
      • 4.3 Dec-POMDP and Features
      • 4.4 Common Knowledge
      • 4.5 Multi-Agent Common Knowledge Reinforcement Learning
      • 4.6 Pairwise MACKRL
      • 4.7 Experiments and Results
      • 4.8 Conclusion & Future Work
    • 5 Stabilizing Experience Replay
      • 5.1 Introduction
      • 5.2 Related Work
      • 5.3 Methods
        • 5.3.1 Multi-Agent Importance Sampling
        • 5.3.2 Multi-Agent Fingerprints
      • 5.4 Experiments
        • 5.4.1 Architecture
      • 5.5 Results
        • 5.5.1 Importance Sampling
        • 5.5.2 Fingerprints
        • 5.5.3 Informative Trajectories
      • 5.6 Conclusion & Future Work
  • II Learning to Communicate
    • 6. Learning to Communicate with Deep Multi-Agent ReinforcementLearning
      • 6.1 Introduction
      • 6.2 Related Work
      • 6.3 Setting
      • 6.4 Methods
        • 6.4.1 Reinforced Inter-Agent Learning
        • 6.4.2 Differentiable Inter-Agent Learning
      • 6.5 DIAL Details
      • 6.6 Experiments
        • 6.6.1 Model Architecture
        • 6.6.2 Switch Riddle
        • 6.6.3 MNIST Games
        • 6.6.4 Effect of Channel Noise
      • 6.7 Conclusion & Future Work
    • 7. Bayesian Action Decoder
      • 7.1 Introduction
      • 7.2 Setting
      • 7.3 Method
        • 7.3.1 Public belief
        • 7.3.2 Public Belief MDP
        • 7.3.3 Sampling Deterministic Partial Policies
        • 7.3.4 Factorized Belief Updates
        • 7.3.5 Self-Consistent Beliefs
      • 7.4 Experiments and Results
        • 7.4.1 Matrix Game
        • 7.4.2 Hanabi
        • 7.4.3 Observations and Actions
        • 7.4.4 Beliefs in Hanabi
        • 7.4.5 Architecture Details for Baselines and Method
        • 7.4.6 Hyperparamters
        • 7.4.7 Results on Hanabi
      • 7.5 Related Work
        • 7.5.1 Learning to Communicate
        • 7.5.2 Research on Hanabi
        • 7.5.3 Belief State Methods
      • 7.6 Conclusion & Future Work
  • III Learning to Reciprocate
    • 8. Learning with Opponent-Learning Awareness
      • 8.1 Introduction
      • 8.2 Related Work
      • 8.3 Methods
        • 8.3.1 Naive Learner
        • 8.3.2 Learning with Opponent Learning Awareness
        • 8.3.3. Learning via Policy gradient
        • 8.3.4 LOLA with Opponent modeling
        • 8.3.5 Higher-Order LOLA
      • 8.4 Experimental Setup
        • 8.4.1 Iterated Games
        • 8.4.2 Coin Game
        • 8.4.3 Training Details
      • 8.5 Results
        • 8.5.1 Iterated Games
        • 8.5.2 Coin Game
        • 8.5.3 Exploitability of LOLA
      • 8.6 Conclusion & Future Work
    • 9. DiCE: The Infinitely Differentiable Monte Carlo Estimator
      • 9.1 Introduction
      • 9.2 Background
        • 9.2.1 Stochastic Computation Graphs
        • 9.2.2 Surrogate Losses
      • 9.3 Higher Order Gradients
        • 9.3.1 Higher Order Gradient Estimators
        • 9.3.2 Higher Order Surrogate Losses
        • 9.3.3. Simple Failing Example
      • 9.4 Correct Gradient Estimators with DiCE
        • 9.4.1 Implement of DiCE
        • 9.4.2 Casuality
        • 9.4.3 First Order Variance Reduction
        • 9.4.4 Hessian-Vector Product
      • 9.5 Case Studies
        • 9.5.1 Empirical Verification
        • 9.5.2 DiCE For multi-agent RL
      • 9.6 Related Work
      • 9.7 Conclusion & Future Work
  • Reference
    • Reference
  • After
    • ๋ณด์ถฉ
    • ์—ญ์ž ํ›„๊ธฐ
Powered by GitBook
On this page
  • Centralized Control
  • Decentralized Control

Was this helpful?

  1. 2. Background
  2. 2. BACKGROUND

2.3 Centralized vs Decentralized Control

Previous2.2 Multi-Agent SettingsNext2.4 Cooperative, Zero-sum, and General-Sum

Last updated 4 years ago

Was this helpful?

Centralized Control

์ „์ฒด๋ฅผ ๊ด€์ธกํ•  ์ˆ˜ ์žˆ๋Š” ์ƒํ™ฉ(fully observable)์—์„œ Multi Agent๋ฅผ ๋งŒ๋“œ๋Š” ๊ฒƒ๋ณด๋‹ค, ์ „์ฒด๋ฅผ ์ด๊ด„ํ•˜๋Š” ํ•˜๋‚˜์˜ Agent(centralized controller) ฯ€C(uโˆฃst)\pi^C(\bold{u}|s_t)ฯ€C(uโˆฃstโ€‹)๋ฅผ ๋งŒ๋“œ๋Š” ๊ฒƒ์„ ์ƒ๊ฐํ•ด ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฅผ ๋‚˜ํƒ€๋‚ด๋ฉด,

ฯ€C(uโˆฃst):Uร—Sโ†’[0,1] \pi^C(\bold{u}|s_t):\bold{U}\times S \rightarrow [0,1] ฯ€C(uโˆฃstโ€‹):Uร—Sโ†’[0,1]

๋กœ ๋‚˜ํƒ€๋‚ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์ด๋Š” ๊ทผ๋ณธ์ ์œผ๋กœ ํฐ ๋ฌธ์ œ์  ๋‘๊ฐ€์ง€๊ฐ€ ์กด์žฌํ•ฉ๋‹ˆ๋‹ค.

  • joint action space U\bold{U}U๋Š” agent๋“ค์˜ action์ด combinatorialํ•˜๊ฒŒ ๊ฒฐํ•ฉ๋œ ํ˜•ํƒœ์ž…๋‹ˆ๋‹ค.

    • P(u1โˆฃs1)โ‹…P(u2โˆฃs2)โ‹…...โ‹…P(unโˆฃsn) P(u^1|s^1)\cdot P(u^2|s^2) \cdot ...\cdot P(u^n|s^n) P(u1โˆฃs1)โ‹…P(u2โˆฃs2)โ‹…...โ‹…P(unโˆฃsn)

    • ์ด๋Š” agent์˜ action space์˜ exponentialํ•œ ์ฆ๊ฐ€๋ฅผ ์˜๋ฏธํ•˜๋ฏ€๋กœ ํ™•์žฅ์„ฑ์—์„œ ๊ต‰์žฅํ•œ ์ œ์•ฝ์ด ๋ฉ๋‹ˆ๋‹ค.

  • local observation ์ƒํ™ฉ์—์„œ์˜ ์ ์šฉ์ด ๋ถˆ๊ฐ€ํ•ฉ๋‹ˆ๋‹ค. ๋งŽ์€ ๋‹ค์–‘ํ•œ ์ƒํ™ฉ์—์„œ agent์˜ observation์€ ์ œํ•œ๋˜๋Š” ์ƒํ™ฉ์ด ์˜ค๋Š”๋ฐ, ์ด๋Š” centralized controller์˜ ์ ์šฉ์ด ๋ถˆ๊ฐ€ํ•จ์„ ๋‚˜ํƒ€๋ƒ…๋‹ˆ๋‹ค.

Decentralized Control

Agent๊ฐ์ž local policy๋ฅผ ์‚ฌ์šฉํ•œ๋‹ค๋ฉด, centralized control์˜ action space์— ๋Œ€ํ•œ ๋‹จ์ ์„ ๊ทน๋ณตํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋•Œ ํ•œ agent์˜ policy๋ฅผ ๋‹ค์Œ๊ณผ ๊ฐ™์ด ํ‘œํ˜„ ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.

ฯ€a(uaโˆฃst)\pi^a(u^a|s_t)ฯ€a(uaโˆฃstโ€‹)

๊ทธ๋ ‡๋‹ค๋ฉด, ์ „์ฒด joint-action์— ๋Œ€ํ•œ ํ™•๋ฅ ์€ ๋‹ค์Œ๊ณผ ๊ฐ™์ด ๋‚˜ํƒ€๋‚ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

P(uโˆฃst)=โˆaฯ€a(uaโˆฃst)P(\bold{u}|s_t) = \prod_a{\pi^a(u^a|s_t)}P(uโˆฃstโ€‹)=โˆaโ€‹ฯ€a(uaโˆฃstโ€‹)