I work on building and understanding intelligent heuristics in the field of machine learning, with special interests on the following topics:
• neurosymbolic visual understanding, generation, and decision making: bridging symbolic (logical, causal) inference and neural network learning to jointly solve system-1 and system-2 visual learning problems.
• active learning, machine teaching, and pretrained model reuse: understanding how prior information can help reduce the sample complexity.
Our small research subgroup is named "future-item". [GitHub]
• For prospective students who are interested to work with me, please read this note first.
Research Papers
Preprints:
• Jingming Liu, Yumeng Li, Wei Shi, Yao-Xiang Ding, Hui Su, Kun Zhou, "Harnessing the Power of Reinforcement Learning for Language-Model-Based Information Retriever via Query-Document Co-Augmentation", arxiv preprint 2506.18670.
• Jingming Liu, Yumeng Li, Boyuan Xiao, Yichang Jian, Ziang Qin, Tianjia Shao, Yao-Xiang Ding, Kun Zhou, "Autonomous Imagination: Closed-Loop Decomposition of Visual-to-Textual Conversion in Visual Reasoning for Multimodal Large Language Models", Transactions on Machine Learning Research (TMLR), 2025. [arXiv] [Project Page]
• Zipei Chen, Yumeng Li, Zhong Ren, Yao-Xiang Ding, Kun Zhou, "Appearance as Reliable Evidence: Reconciling Appearance and Generative Priors for Monocular Motion Estimation", Computers & Graphics (CAD/Graphics 2025 conference best paper award), 2025.
• Yuting Tang, Xin-Qiang Cai, Yao-Xiang Ding, Qiyu Wu, Guoqing Liu, Masashi Sugiyama, "Reinforcement Learning from Bagged Reward", Transactions on Machine Learning Research (TMLR), 2025. [arXiv]
• Lanjihong Ma, Yao-Xiang Ding, Peng Zhao, Zhi-Hua Zhou, "Learning Objective Adaptation by Correlation-based Model Reuse", IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2025.
• Yumeng Li, Bohong Chen, Zhong Ren, Yao-Xiang Ding, Libin Liu, Tianjia Shao, Kun Zhou, "CPoser: An Optimization-after-Parsing Approach for Text-to-Pose
Generation Using Large Language Models", ACM Transactions on Graphics (TOG) (SIGGRAPH Asia journal track paper), 2024.
• Yu-Cheng He, Yao-Xiang Ding, Zhi-Hua Zhou, "Mechanism Design for Requester-Platform Strategies Under the Three-Party Crowdsourcing Market", Journal of Computer Research And Development (in Chinese, CCFAI 2021 conference best paper award), 2021.
• Dingsheng Luo, Yaoxiang Ding, Xiaoqiang Han, Yang Ma, Yian Deng, Zhan Liu, Xihong Wu, "Humanoid Environmental Perception with Gaussian Process Regression", International Journal of Advanced Robotic Systems, 2016.
Activities
•
Model Reuse: Theories and Applications, Tutorial on AAAI'24, co-organized with Dr. Han-Jia Ye. [Slides of part1][Slides of part2]
Courses
• Advanced Data Structures and Algorithm Analysis (page under construction), Undergraduate, Fall and Winter 2025, Zhejiang University. Previous versions: [Fall and Winter 2024][Spring and Summer 2024]
• Introduction to Machine Learning (co-teach with Prof. Yingchun Yang, Dr. Qian Zheng and Dr. Chaochao Chen), Graduate, Spring 2023, Zhejiang University. [Slides of Lecture 7][Slides of Lecture 8]