WCCI 2022 Special Session

Special Session: Data-Efficient Multi-Agent Reinforcement Learning for Control, Automation, and Population Systems Optimization


Large populations of isolated or interacting dynamical systems are ubiquitous in diverse disciplines of science and engineering with applications ranging from robotic swarms, smart-grid, and wireless sensor networks to neural systems, opinion networks, and cellular oscillators. This special session aims to provide a forum to encourage interaction among researchers and partitioners in reinforcement learning (RL), adaptive/approximate dynamic programming (ADP), and computational intelligence (CI). This special session welcomes the contributions from multiple research fields, including control theory, computer science, computational and artificial intelligence, and so on, to deliver and discuss the frontier research results and new techniques of data-efficient multi-agent RL for control and optimization. Data-efficient multi-agent RL is becoming increasingly relevant to handle sequential and continuous decision-making problems in control, automation, and population system optimization due to its unique scalability, practicality, and optimality. Data-efficient multi-agent RL has been demonstrated with incredible success and will be the key technique to a wide range of emerging applications, such as autonomous vehicles, intelligent robots, smart-grid, smart communities, and so on. Computational intelligence techniques, including biological inspired neural networks, fuzzy logic, and evolution computation, have the values to be integrated into the multi-agent interaction process to deal with the important research problems of learning, approximation, and generalization.

Scope and Topics:

This special session will provide a forum to deliver and discuss original research results and new techniques in deep, efficient RL for large-scale multi-agent systems. We are particularly interested in the following topics:

  • Deep efficient RL design for large scale multi-agent systems
  • Deep hybrid RL based optimal control
  • Deep safe RL based robust adaptive control
  • Deep RL based event-triggered/self-triggered control
  • Biologically inspired deep RL for multiplayer games
  • Novel Deep efficient RL algorithms, stability analysis, and convergence
  • Meta-learning and Transfer learning for autonomous systems
  • Applications of large scale multi-agent RL and ADP (e.g., autonomous vehicles, smart grid, intelligent robots, and others)
  • Multi-agent learning rules and theories
  • Large scale multi-agent game and applications
  • Partially observable multi-agent learning
  • Bayesian methods for multi-agent learning
  • Distributed multi-agent interactions
  • Multi-objective multi-agent methods
  • Transfer learning in multi-agent problems

Important Dates:

  • Paper Submission Deadline: January 31, 2022
  • Notification of Acceptance: April 26, 2022
  • Final Paper Submission Deadline: May 23, 2022

Call for paper link: https://wcci2022.org/call-for-papers/

Pape Submission:

Name of Organizers:

Zhen Ni, Florida Atlantic University, FL, USA,

Xiangnan Zhong, Florida Atlantic University, FL, USA,

Avimanyu Sahoo, Oklahoma State University, OK, USA, 

Vignesh Narayanan, University of Southern Carolina, SC, USA, 

Hao Xu, University of Nevada, NV, USA,