Ø 个人简介

郭泓树，博士，华南师范大学人工智能学院特聘研究员，研究基于先进机器学习方法的黑箱优化算法自动设计，探索通过深度强化学习、大语言模型等先进学习方法赋予黑箱优化算法学习与泛化能力，实现黑箱优化算法的自主设计与自适应配置。在深度强化学习、演化计算和大语言模型等领域有丰富的工作经验，已在国际高水平学术期刊或会议投稿和发表论文24篇，其中第一作者（含共一）5篇，google scholor h-index 为12，i10-index为14，总引用量486（截止2026-06）。本人担任多个顶尖国际期刊和会议的审稿人，包括NuerIPS、ICLR、ICML等国际先进学术会议和IEEE TEVC等SCI Q1期刊。欢迎对相关领域有兴趣的同学联系交流（guohongshu369@gmail.com）。

Ø 教育背景

2022年9月 ~ 2026年6月：  华南理工大学，工学博士学位，导师：龚月姣教授（广东省杰青，IEEE TEVC副主编，Stanford全球前2%科学家）；

2018年9月 ~ 2022年6月：  华南理工大学，学士学位；

Ø 工作经历

2026年7月 ~ 至今：  加入华南师范大学人工智能学院，特聘研究员；

Ø 研究领域

进化计算 , 机器学习 , 深度强化学习 , 黑箱优化 , 大语言模型

Ø 代表性论文

  期刊论文

[1] Ma Zeyuan, Gong Yue-Jiao*, Guo Hongshu, Chen Jiacheng, Ma Yining, Cao Zhiguang, Zhang Jun. Llamoco: Instruction tuning of large language models for optimization code generation[J]. IEEE Transactions on Evolutionary Computation, 2026.

[2] Ma Zeyuan, Guo Hongshu, Gong Yue-Jiao*, Zhang Jun, Tan Kay Chen. Toward automated algorithm design: A survey and practical guide to meta-black-box-optimization[J]. IEEE Transactions on Evolutionary Computation, 2025.

[3] Guo Hongshu, Ma Yining, Ma Zeyuan, Chen Jiacheng, Zhang Xinglin, Cao Zhiguang, Zhang Jun, Gong Yue-Jiao*. Deep reinforcement learning for dynamic algorithm selection: A proof-of-principle study on differential evolution[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2024.

  会议论文

[1] Ma Zeyuan, Cao Zhiguang, Jiang Zhou, Guo Hongshu, Gong Yue-Jiao*. Meta-Black-Box-Optimization through Offline Q-function Learning[C]. Forty-second International Conference on Machine Learning , 2026.

[2] Gao Zijian, Zhong Yuanting, Ma Zeyuan, Gong Yue-Jiao, Guo Hongshu*. Detect and Act: Automated Dynamic Optimizer through Meta-Black-Box Optimization[C]. Proceedings of the Genetic and Evolutionary Computation Conference Companion, 2026.07.

[3] Ma Zeyuan, Gong Yue-Jiao*, Guo Hongshu, Qiu Wenjie, Ma Sijie, Lian Hongqiao, Zhan Jiajun, Chen Kaixu, Wang Chen, Huang Zhiyang, others. Metabox-v2: A unified benchmark platform for meta-black-box optimization[C]. Advances in Neural Information Processing Systems, 2026.01.

[4] Guo Hongshu, Ma Zeyuan, Ma Yining, Zhang Xinglin, Chen Wei-Neng, Gong Yue-Jiao*. Designx: Human-competitive algorithm designer for black-box optimization[C]. Advances in Neural Information Processing Systems, 2026.01.

[5] Chen Jiacheng, Ma Zeyuan, Guo Hongshu, Ma Yining, Zhang Jie, Gong Yue-Jiao*. SYMBOL: Generating Flexible Black-Box Optimizers through Symbolic Equation Learning[C]. The Twelfth International Conference on Learning Representations , 2024.

Ø 联系方式

guohongshu369@gmail.com