Publications

You can also find my articles on my Google Scholar. My research can be summarised by the following keywords:

• Methodologies-Oriented: Generative Models; Flow Matching; Diffusion Models; Unsupervised Machine Learning; Anomaly Detection; Transfer Learning; XAI; Graph Neural Networks
• Applications-Oriented: Log Analysis; Predictive Maintenance; Digital Twin; AI for Manufacturing

2026

• MM-OptBench: A Solver-Grounded Benchmark for Multimodal Optimization Modeling.
  • Zhong Li, Qi Huang, Yuxuan Zhu, Mohammad Mohammadi Amiri, Niki van Stein, Thomas Bäck, Matthijs van Leeuwen, Zaiwen Wen, Lincen Yang. Under Review for an AI conference. “#” means equal contribution
• OptArgus: A Multi-Agent System to Detect Hallucinations in LLM-based Optimization Modeling.
  • Zhong Li, Zihan Guo, Xiaohan Lu, Juntao Wang, Jie Song, Chao Shen, Jiageng Wu, Mingyang Sun. Under Review for an AI conference. “#” means equal contribution
• Latent Space Analaysis for Unsupervised Hyperparameter Optimisation.
  • Yuhang Wang #, Zhong Li #, Shujian Yu, Matthijs van Leeuwen. Under Review for an AI conference. “#” means equal contribution
• Diffusion and Flow Matching Models for Tabular Data: A Survey.
  • Zhong Li, Qi Huang, Lincen Yang*, Jiayang Shi, Zhao Yang, Niki van Stein, Thomas Bäck, Matthijs van Leeuwen. Under Review for TKDE. “*” means corresponding author
  • [paper], [code]
• Explainable Graph Neural Networks under Fire.
  • Zhong Li, Simon Geisler, Yuhang Wang, Stephan Günnemann, Matthijs van Leeuwen. Under Review for an AI conference.
  • [paper], [code]
• Constructing Industrial-Scale Optimization Modeling Benchmark.
  • Zhong Li, Hongliang Lu, Tao Wei, Yuxuan Chen, Wenyu Liu, Yuan Lan, Fan Zhang, Zaiwen Wen. ICML 2026.
  • [paper], [code]

2025

• Learning Subgroups with Maximum Treatment Effects without Causal Heuristics.
  • Lincen Yang, Zhong Li*, Matthijs van Leeuwen, Saber Salehkaleybar. AAAI 2026 Oral. “*” means corresponding author.
  • [paper], [code]
• Scalable, Explainable and Provably Robust Anomaly Detection with One-Step Flow Matching.
  • Zhong Li, Qi Huang, Yuxuan Zhu, Lincen Yang*, Mohammad Mohammadi Amiri, Niki van Stein, Matthijs van Leeuwen. NeurIPS 2025. “*” means corresponding author.
  • [paper], [code]
• Towards Automated Self-Supervised Learning for Truly Unsupervised Graph Anomaly Detection.
  • Zhong Li, Yuhang Wang, Matthijs van Leeuwen. Data Mining and Knowledge Discovery (DAMI). 39(44): 1-43. (AAAI 2026 Journal Track)
  • [paper], [code]

2024

• Graph Neural Networks based Log Anomaly Detection and Explanation.
  • Zhong Li, Jiayang Shi, Matthijs van Leeuwen. ICSE’24 poster track.
  • [paper], [code] (ArXiv full version)
  • [paper] (ICSE short version)
• Cross-Domain Graph Level Anomaly Detection.
  • Zhong Li, Sheng Liang, Jiayang Shi, Matthijs van Leeuwen. IEEE Transactions on Knowledge and Data Engineering (TKDE). 36(12):7839-7850
  • [paper], [code]

2023

• A Survey on Explainable Anomaly Detection.
  • Zhong Li, Yuxuan Zhu, Matthijs van Leeuwen. ACM Transactions on Knowledge Discovery from Data (TKDD). 23: 1–54.
  • [paper]
• Explainable Contextual Anomaly Detection using Quantile Regression Forests.
  • Zhong Li, Matthijs van Leeuwen. Data Mining and Knowledge Discovery (DAMI). 37: 2517–2563.
  • [paper], [code]

2022

• Feature Selection for Fault Detection and Prediction based on Log Analysis.
  • Zhong Li, Matthijs van Leeuwen. SIGKDD Explorations Newsletter. 24(2): 96–104.
  • [paper]
• Choosing and changing the analysis scale in non-inferiority trials with a binary outcome.
  • Zhong Li, Matteo Quartagno, Stefan Böhringer, and Nan van Geloven. Clinical Trials. 19(1): 14-21.
  • [paper]