Dr. Xudong Wang
I received my PhD in Computer Science from the School of Data Science, CUHK-Shenzhen. My research focuses on artificial intelligence and machine learning, especially graph learning, smart grid intelligence, LLMs, multimodal learning, and energy analytics. My CUHK-Shenzhen profile is available here, and I can be reached at xudongwang@link.cuhk.edu.cn.
My PhD research was primarily supervised by Prof. Chris Ding (丁宏强) and co-supervised by Prof. Tongxin Li (李彤欣). During my PhD, I also worked closely with Prof. Jicong Fan (樊继聪) and Prof. Guoming Tang (唐国明), and I am grateful for their research guidance and collaboration.
My work has appeared in leading machine learning and energy-system venues, including ICML, AAAI, IJCAI, ACM e-Energy, and ACM BuildSys, with presentations across these communities. I received the IEEE SustainCom 2024 Best Paper Award and the CUHK-Shenzhen Presidential Award for Outstanding Graduate Students. I regularly serve as a reviewer for leading AI and smart grid conferences and journals, and served as a Session Chair at PAKDD 2026. My current interests include LLM-enabled foundation models and multi-agent systems for graph learning and smart grid intelligence.
Before my PhD, I received an M.Sc. in Data Science from CUHK-Shenzhen in 2022, advised by Prof. Chris Ding. I received a B.Sc. in Statistics from the School of Mathematics, Shandong University, and a B.Econ. in Finance from the School of Economics, Shandong University, in 2020.
Latest News Recent updates News and selected updates.
🇰🇷 I will attend ICML 2026 in Seoul, South Korea from July 6-11. Poster: Learnable Kernel Density Estimation for Graphs and Its Application to Graph-Level Anomaly Detection, Thu, Jul 9, 2026, 3:30 AM-5:15 AM CEST, Hall A #2403. Please stop by for discussion or a coffee chat.
⚡ Our ACM e-Energy 2026 paper, Energy Injection Identification enabled Disaggregation with Deep Multi-Task Learning, was presented in Banff, Canada. Due to visa constraints, the presentation was delivered by Prof. S. Keshav.
📚 At PAKDD 2026 in Hong Kong, I co-presented Tutorial 3: Foundations and Recent Advances of Graph Learning: A Perspective from Distances and Representations. I also chaired Session 2D, Distribution Shift and Evaluation, on June 10, 2:00-3:00pm, Room 4.
🎓 I passed my PhD defense. I expect to formally receive the CUHK CS PhD degree in October 2026. I am open to opportunities in AI algorithm research, quantitative trading, and related roles; happy to connect.
Research Research Focus Principled graph learning, smart grid AI, LLMs, multimodal learning, and efficient ML systems.
Principled Graph Learning
Principled graph representation learning built around informative unsupervised objectives, interpretable structural patterns, geometry-aware embeddings, distribution modeling, and graph foundation models.
Smart Grid AI
Robust AI for real-world energy systems, including non-intrusive load monitoring, EV charging detection, energy injection identification, and LLM-enhanced decision making.
LLMs, Agents, and Efficient ML
LLM-enabled energy analytics, domain foundation models, multi-agent reasoning, and memory-efficient long-context learning systems.
王旭东 博士
我是香港中文大学（深圳）数据科学学院计算机科学博士。我的研究方向是人工智能与机器学习，重点关注图学习、智能电网、大语言模型、多模态学习，以及面向能源系统的数据分析方法。我在香港中文大学（深圳）的个人主页见这里，也欢迎通过 xudongwang@link.cuhk.edu.cn 联系我。
我的博士研究主要由丁宏强教授（Prof. Chris Ding）指导，并由李彤欣教授（Prof. Tongxin Li）联合指导；硕士阶段也师从丁宏强教授。博士期间，我还与樊继聪教授和唐国明教授紧密合作，并有幸得到他们在科研上的指导与帮助。
我的工作发表于 ICML、AAAI、IJCAI 等机器学习与人工智能领域顶级会议，以及 ACM e-Energy、ACM BuildSys 等智能能源系统会议，并在相关会议进行报告与展示；曾获 IEEE SustainCom 2024 Best Paper Award、香港中文大学（深圳）校长杰出研究生奖等荣誉。我长期担任人工智能领域国际会议和期刊审稿人，并曾担任 PAKDD 2026 Session Chair。目前，我也关注结合 LLMs 的图学习与能源系统基础模型、多智能体方法等方向。
此前，我于 2022 年在香港中文大学（深圳）获得数据科学硕士学位。本科阶段，我于 2020 年获山东大学数学学院统计学 B.Sc. 学位，并获山东大学经济学院金融学 B.Econ. 辅修双学位。
最新动态 近期动态 新闻与近期更新。
🇰🇷 我将于 2026.07.06-07.11 在韩国首尔参加 ICML 2026。Poster：Learnable Kernel Density Estimation for Graphs and Its Application to Graph-Level Anomaly Detection，时间为 Thu, Jul 9, 2026, 3:30 AM-5:15 AM CEST，地点 Hall A #2403，欢迎来交流或 coffee chat。
⚡ 我们的 ACM e-Energy 2026 论文 Energy Injection Identification enabled Disaggregation with Deep Multi-Task Learning 已在加拿大 Banff 展示。由于签证原因，本次报告由 Prof. S. Keshav 代为 Presentation。
📚 在香港 PAKDD 2026，我共同主讲 Tutorial 3：Foundations and Recent Advances of Graph Learning: A Perspective from Distances and Representations。同时担任 Session 2D, Distribution Shift and Evaluation 的 Session Chair，时间为 2026.06.10 2:00-3:00pm，地点 Room 4。
🎓 博士毕业答辩已通过，预计将于 2026 年 10 月正式取得香港中文大学（CUHK）计算机科学博士学位。目前 open for job，关注人工智能算法研究、量化交易等方向，欢迎交流。
研究方向 研究聚焦 Principled graph learning, smart grid AI, LLMs, multimodal learning, and efficient ML systems.
理论驱动的图学习
围绕高效无监督目标、可解释结构模式、几何自适应表示、图分布建模，以及图基础模型，构建理论驱动且面向应用的图表示学习方法。
智能电网 AI
面向真实能源系统的稳健人工智能方法，包括非侵入式负荷监测、电动汽车充电检测、能量注入识别，以及 LLM 增强的能源决策。
LLMs、智能体与高效机器学习
面向能源分析的大语言模型、领域基础模型、多智能体推理，以及长上下文场景下的高效学习系统。