Our Research

Advancing superintelligence (SI) through research and innovation

Core Research Pillars

At the intersection of theoretical innovation and transformative industrial applications

Deep Learning & Neural Architectures

We pioneer advanced deep learning architectures including Graph Neural Networks (GNNs), Transformers, and Recurrent Attention Models for complex data analysis. Our innovations include spatiotemporal graph networks for cellular KPI prediction, multi-task learning frameworks for medical imaging, and deep reinforcement learning for bilevel optimization in product design.

Statistical Learning & Monitoring for SI

Combining statistics and machine learning, we develop models for real-time monitoring, anomaly detection, and predictive maintenance. Our methods include Bayesian networks for causal inference, Gaussian processes for robust parameter design, and Thompson sampling for partially observable online change detection.

Tensor Learning & High-Dimensional Analytics

We tackle the curse of dimensionality through innovative tensor methods and functional data analysis. Our research includes tensor decomposition for weakly-dependent data on graphs, low-rank robust subspace tensor clustering for metro flow modeling, and latent tensor Gaussian processes for spatial monitoring.

Graph Intelligence & Network Analysis

Advancing graph relational learning and network science for SI. We develop graph regularized tensor latent Dirichlet allocation for individualized passenger travel patterns, functional data edged networks for metro station clustering, and multi-view clustering based on functional data.

Intelligent Decision Systems & Optimization

Creating decision frameworks for SI that integrate game theory, optimization, and reinforcement learning. Our research spans deep reinforcement learning for bilevel optimization, data-driven control of complex networked systems, and intelligent decision-making with domain knowledge integration.

Industrial Superintelligence & Smart Manufacturing

Developing practical methods for industrial superintelligence. Our applications span semiconductor manufacturing, aircraft manufacturing, urban transportation, supply chain management, and financial systems. We apply our research at scale with industry leaders including Huawei, Tencent, Alibaba, Baidu, ByteDance, and Meituan.

Research Impact

Our research has resulted in 81 publications in premier venues including IEEE Transactions (Medical Imaging, Knowledge and Data Engineering, Automation Science and Engineering, Industrial Informatics), IISE Transactions, Technometrics, Journal of Quality Technology, and leading research conferences including KDD, IJCAI, AAAI, and ECML-PKDD.

We collaborate extensively with industry leaders, developing methods that support users daily in transportation, manufacturing, e-commerce, and telecommunications sectors.

Recent Publications

Highlights from our latest research contributions

A Computer-Aided System for Enhanced M-protein Diagnosis and Monitoring via Multimodal Learning
An, X.; Li, P.*; Han, J.; Ji, W.; Di, Q.; Liang, Y.; Hou, Y.; Sha, Y.; Su, W.; Qiu, L.; Zhang, C.*
2026 · Nature Communications — accepted
Paper link forthcoming
VETime: Vision Enhanced Zero-Shot Time Series Anomaly Detection
Yingyuan Yang, Tian Lan, Yifei Gao, Yimeng Lu, Wenjun He, Meng Wang, Chenghao Liu, Chen Zhang
2026 · NeurIPS 2026 — accepted
Supervision Recovery for Time Series Anomaly Detection via Counterfactual Pairing
Yifei Gao, Tian Lan, Yimeng Lu, Xuming An, Meng Wang, Wenjun He, Yijie Li, Chen Zhang
2026 · NeurIPS 2026 — accepted
Preprint title: Supervision Recovery for Time Series Anomaly Detection via Context-Anchored Pairing.
Learning Collision-Free Dispatch Policies for Route-Wise Decision-Dependent Anomaly Detection
Xa, X.; Zhang, C.; Liu, K.*
2026 · NeurIPS 2026 — accepted
Quickest Causal Change Point Detection by Adaptive Intervention
Haijie Xu, Chen Zhang
2026 · IEEE Transactions on Pattern Analysis and Machine Intelligence — accepted
SCALE: Style-Causal Disentanglement with Adaptive Lifelong Expert for Online Latent-domain Anomaly Detection
Yimeng Lu, Yifei Gao, Tian Lan, Yingyuan Yang, Wenjun He, Meng Wang, Chen Zhang
2026 · Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2, 3436-3447
View All Publications (81)