Skip to content
@GMC-DRL

GMC-Research-Team

We are a research team mainly focus on Meta-Black-Box Optimization.

🌟 Welcome to GMC team 🌟

😎 Who is Us

We are a research team mainly focus on the Evolutionary Computing, Deep Reinforcement Learning, Black Box Optimization and Meta Black Box Optimization. We belong to the Computational Intelligence Lab, School of Computer Science, South China University of Technology. (Our homepage)

We are an energetic team including undergraduate students, master students and phd students. The student leader in this team is Zeyuan Ma, a phd student at South China University of Technology. Students in GMC team are advised (in part advised) by Prof. Yue-Jiao Gong. This is a pure research-oriented technical team, aiming to develop the new generation of black-box-optimization concepts, algorithms, frameworks and benchmarks. The resulting research domain is commonly named as Meta-Black-Box-Optimization, which generally mitigates the labour-intensive development in traditional black-box optimization algorithms through meta-learning an update rule/algorithmic configuration at the meta level. We believe works done in this team would promote the study edge of both evolutionary computing and optimization.


Advisor

Prof. Yue-Jiao Gong

Student Founder

Zeyuan Ma (PhD, 22-26)

Hongshu Guo (PhD, 22-26)

Student Member

Undergraduate

Name Year in Group Post-Graduation Career
Jiacheng Chen(*co-founder) 2021 - -
Zhenrui Li 2021 - 2023 -
Kaixu Chen 2023 - -
Wenjie Qiu 2023 - -
Sijie Ma 2023 - -
Yuzhi Hu 2023 - -
Zechuan Huang 2023 - -

Master

Name Year in Group Post-Graduation Career
Hongqiao Lian 2023 - -
Jiajun Zhan 2023 - -
Guojun Peng 2023 - -
Jianhao Liu 2022 - -

Cooperator

Prof. Zhiguang Cao

Yining Ma


🚩 What Have We Done

Benchmarking MetaBBO Approaches

Year Paper
2023 Zeyuan Ma, et al. "MetaBox: A Benchmark Platform for Meta-Black-Box Optimization with Reinforcement Learning." Advances in Neural Information Processing Systems 36 (NeurIPS 2023).

Developing MetaBBO Approaches

Year Paper
2024 Zeyuan Ma*, Jiacheng Chen*, Hongshu Guo, et. al. "Auto-configuring Exploration-Exploitation Tradeoff in Evolutionary Computation via Deep Reinforcement Learning" The Genetic and Evolutionary Computation Conference (2024). to be published.
2024 Hongqiao Lian, Zeyuan Ma, Hongshu Guo, et. al. "RLEMMO: Evolutionary Multimodal Optimization Assisted By Deep Reinforcement Learning" The Genetic and Evolutionary Computation Conference (2024). to be published.
2024 Hongshu Guo, Zeyuan Ma, Jiacheng Chen, et. al. "Deep Reinforcement Learning for Dynamic Algorithm Selection: A Proof-of-Principle Study on Differential Evolution" IEEE Transactions on Systems, Man, and Cybernetics: Systems (2024).
2024 Jiacheng Chen*, Zeyuan Ma*, et. al. "Symbol: Generating Flexible Black-Box Optimizers through Symbolic Equation Learning" The Twelfth International Conference on Learning Representations (ICLR 2024)

Exploring AGI (AIGC) in Optimization

Year Paper
2024 Zeyuan Ma, Hongshu Guo, Jiacheng Chen, et. al. "LLaMoCo: Instruction Tuning of Large Language Models for Optimization Code Generation"

Other Topics in Optimization

Year Paper
2023 SC Lei, Hongshu Guo, et. al. "A High-Performance Tensorial Evolutionary Computation for Solving Spatial Optimization Problems" International Conference on Neural Information Processing (ICONIP 2023)
2021 Zeyuan Ma, et. al. "An efficient computational approach for automatic itinerary planning on web servers" Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2021)

⚡ What is Coming

Student Exchange

Name Time Destination
Jiacheng Chen 2024/06 - 2024/09 Computing + Mathematical Sciences (CMS) Department, Caltech

Recent activities

Activity Type Title Date
AITIME NeurIPS 2023 Pre-Talk talk MetaBox: 面向元黑箱优化兼容的黑箱优化测试平台 2023/11
AITIME ICLR 2024 Pre-Talk talk Symbol: Generating Flexible Black-Box Optimizers through Symbolic Equation Learning 2024/03

🏠 Our Repositories

  • MetaBox: The first benchmark platform expressly tailored for developing and evaluating MetaBBO-RL methods, which is accepted at NeurIPS 2023.
  • psc4MetaBBO: A list of useful relevant papers and open source codes for MetaBBO.
  • Symbol: The python implementation of our paper SYMBOL, which is accepted as a poster paper at ICLR 2024. This is a novel MetaBBO paragidm against the recent proposed ones, refer to the paper for detail.
  • Coming soon

📧 Contact Us

You can reach out to ask questions or just chat about us! We are available on E-mail: [email protected]

Pinned Loading

  1. MetaBox MetaBox Public

    MetaBox: A Benchmark Platform for Meta-Black-Box Optimization with Reinforcement Learning (https://arxiv.org/abs/2310.08252)

    Python 54 8

  2. Symbol Symbol Public

    Python implementation of SYMBOL

    Python 11 3

  3. Awesome-MetaBBO Awesome-MetaBBO Public

    A collection of MetaBBO papers and code sources

    13

  4. RL-DAS RL-DAS Public

    Python 1

  5. GLEET GLEET Public

    Python implementation of Auto-configuring Exploration-Exploitation Tradeoff in Evolutionary Computation via Deep Reinforcement Learning

    Python 1

Repositories

Showing 10 of 13 repositories