Author name cluster

Kun Chen

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

17 papers

2 author rows

AAAI Conference 2026 Conference Paper

FedSkeleton: Secure Multi-Party Graph Skeleton Construction for Privacy-Preserving Federated Time-Series Forecasting

Henggang Deng
Yuchao Tang
Wenjie Fu
Huandong Wang
Kun Chen
Tao Jiang

In real-world time-series modelling, graph structures are widely adopted because they explicitly encode node topology and capture complex network dynamics. In practice, however, a complete graph is often partitioned across multiple parties; each party can access only its local sub-graph and, owing to privacy regulations, cannot share topology or data, creating pervasive data silos. Federated Graph Learning (FGL) offers a privacy-preserving collaborative-learning paradigm, yet current methods still face two key challenges: (1) the graph topology itself contains sensitive structural information, which can lead to privacy leakage if directly shared during FGL; (2) cross-party edges are crucial for accurate modeling, yet exploiting them without compromising privacy remains a significant challenge. To overcome these obstacles, we propose FedSkeleton, a privacy-preserving framework for time-series prediction that comprises a Skeleton Construction Module and a Dual-stream Forecasting Module, enabling global dependency capture without revealing the topology. Extensive experiments show that FedSkeleton consistently outperforms existing baselines and even surpasses models trained in a centralized setting with full-graph access in certain cases. In addition, we conduct comprehensive security analysis, communication-cost evaluation and scalability experiments, demonstrating that FedSkeleton effectively resists common attacks, keeps communication overhead manageable, and remains robust with respect to key hyper-parameters and the number of participating parties.

PDF Details DOI

EAAI Journal 2025 Journal Article

Adaptive attention graph convolution network with normalized embedded Gaussian for rapid serial visualization presentation decoding

Mengyuan Zhao
Qingsong Ai
Kun Chen
Quan Liu
Sheng Quan Xie
Li Ma

Graph convolutional networks (GCNs) have been widely used in Brain Computer Interface (BCI) and have shown great prowess in identifying electroencephalogram (EEG) spatiotemporal features. However, most GCNs learn channels topological relationship by fixed adjacency matrix. This lacks connectivity strength information and ignores the data dependency. This paper proposes a data-driven adjacency matrix based on normalized embedded Gaussian function, and constructs a Gaussian-Adaptive Attention Graph Convolution Network (Gaussian-AAGCN). Brain regions connectivity is calculated by normalized embedded Gaussian function, and the topological relationship is adaptively learned by input data in a data-driven manner. This data-driven adaptive adjacency matrix avoids brain activity information loss caused by fixed adjacency matrix and improves the flexibility of graph construction. Convolutional block attention module (CBAM) is introduced to adaptive feature refinement in two independent dimensions, improving model representation ability. Experimental results show that the average area under curve (AUC), true positive rate (TPR) and false positive rate (FPR) of Gaussian-AAGCN on 14 subjects are 93. 52 %, 91. 59 %, and 4. 58 % respectively. Compared to Transformer, Event-Related Potential Capsule Network (ERP-CapsNet), Electroencephalogram Convolutional Neural Network (EEGNet), and Multi-Granularity Information Fusion Network (MGIFNet), the AUC of Gaussian-AAGCN is higher by 18. 02 %, 5. 32 %, 2. 62 %, and 1. 82 %, respectively. Using the adaptive adjacency matrix, the model AUC and TPR are increased by about 4. 8 % and 7. 8 % respectively. After integrating CBAM, the AUC and TPR increased by about 3. 5 % and 8 % respectively.

Details DOI

NeurIPS Conference 2025 Conference Paper

Align-DA: Align Score-based Atmospheric Data Assimilation with Multiple Preferences

Jing-An Sun
Hang Fan
Junchao Gong
Ben Fei
Kun Chen
Fenghua Ling
Wenlong Zhang
Wanghan Xu

Data assimilation (DA) aims to estimate the full state of a dynamical system by combining partial and noisy observations with a prior model forecast, commonly referred to as the background. In atmospheric applications, this problem is fundamentally ill-posed due to the sparsity of observations relative to the high-dimensional state space. Traditional methods address this challenge by simplifying background priors to regularize the solution, which are empirical and require continual tuning for application. Inspired by alignment techniques in text-to-image diffusion models, we propose Align-DA, which formulates DA as a generative process and uses reward signals to guide background priors—replacing manual tuning with data-driven alignment. Specifically, we train a score-based model in the latent space to approximate the background-conditioned prior, and align it using three complementary reward signals for DA: (1) assimilation accuracy, (2) forecast skill initialized from the assimilated state, and (3) physical adherence of the analysis fields. Experiments with multiple reward signals demonstrate consistent improvements in analysis quality across different evaluation metrics and observation-guidance strategies. These results show that preference alignment, implemented as a soft constraint, can automatically adapt complex background priors tailored to DA, offering a promising new direction for advancing the field.

PDF Details

NeurIPS Conference 2025 Conference Paper

DAWP: A framework for global observation forecasting via Data Assimilation and Weather Prediction in satellite observation space

Junchao Gong
Jingyi Xu
Ben Fei
Fenghua Ling
Wenlong Zhang
Kun Chen
Wanghan Xu
Weidong Yang

Weather prediction is a critical task for human society, where impressive progress has been made by training artificial intelligence weather prediction (AIWP) methods with reanalysis data. However, reliance on reanalysis data limits the AIWPs with shortcomings, including data assimilation biases and temporal discrepancies. To liberate AIWPs from the reanalysis data, observation forecasting emerges as a transformative paradigm for weather prediction. One of the key challenges in observation forecasting is learning spatiotemporal dynamics across disparate measurement systems with irregular high-resolution observation data, which constrains the design and prediction of AIWPs. To this end, we propose our DAWP as an innovative framework to enable AIWPs to operate in a complete observation space by initialization with an artificial intelligence data assimilation (AIDA) module. Specifically, our AIDA module applies a mask multi-modality autoencoder (MMAE) for assimilating irregular satellite observation tokens encoded by mask ViT-VAEs. For AIWP, we introduce a spatiotemporal decoupling transformer with cross-regional boundary conditioning (CBC), learning the dynamics in observation space, to enable sub-image-based global observation forecasting. Comprehensive experiments demonstrate that AIDA initialization significantly improves the roll-out and efficiency of AIWP. Additionally, we show that DAWP holds promising potential to be applied in global precipitation forecasting.

PDF Details

NeurIPS Conference 2025 Conference Paper

LoRA-EnVar: Parameter-Efficient Hybrid Ensemble Variational Assimilation for Weather Forecasting

Yi Xiao
Hang Fan
Kun Chen
Ye Cao
Ben Fei
Wei Xue
Lei Bai

Accurate estimation of background error (i. e. , forecast error) distribution is critical for effective data assimilation (DA) in numerical weather prediction (NWP). In state-of-the-art operational DA systems, it is common to account for the temporal evolution of background errors by employing hybrid methods, which blend a static climatological covariance with a flow-dependent ensemble-derived component. While effective to some extent, these methods typically assume Gaussian-distributed errors and rely heavily on hand-crafted covariance structures and domain expertise, limiting their ability to capture the complex, non-Gaussian nature of atmospheric dynamics. In this work, we propose LoRA-EnVar, a novel hybrid ensemble variational DA algorithm that integrates low-rank adaptation (LoRA) into a deep generative modeling framework. We first learn a climatological background error distribution using a variational autoencoder (VAE) trained on historical data. To incorporate flow-dependent uncertainty, we introduce LoRA modules that efficiently adapt the learned distribution in response to flow-dependent ensemble perturbations. Our approach supports online finetuning, enabling dynamic updates of the background error distribution without catastrophic forgetting. We validate LoRA-EnVar in high-resolution assimilation settings using the FengWu forecast model and simulated observations from ERA5 reanalysis. Experimental results show that LoRA-EnVar significantly improves assimilation accuracy over models assuming static background error distribution and achieves comparable or better performance than full finetuning while reducing the number of trainable parameters by three orders of magnitude. This demonstrates the potential of parameter-efficient adaptation for scalable, non-Gaussian DA in operational meteorology.

PDF Details

EAAI Journal 2025 Journal Article

Physics-guided neural network for predicting international roughness index on flexible pavements considering accuracy, uncertainty and stability

Kun Chen
Mehran Eskandari Torbaghan
Nick Thom
Asaad Faramarzi

An outstanding amount of funds are allocated to maintain road network conditions. To ensure the serviceability of roads, the accurate prediction of its roughness or International Roughness Index plays a pivotal role in road management. Artificial neural network, typically used in roughness prediction, is a powerful machine learning algorithm that learns complex patterns in data with non-linear relationships. However, it remains a black box solution and relies heavily on the utilized data and its internal structures, causing model's overfitting and instability. To address such issues, a physics-guided neural network modelling framework is proposed for short- and long-term predictions of roughness aimed at improving model's accuracy, uncertainty and stability. This framework fuses the output of physics-based model simulations along with field observational data acquired from the Long-Term Pavement Performance public database as inputs to develop a neural network architecture. Additionally, the framework uses a physics-based loss function in the network's learning process to ensure the predictions are consistent with the known physics. The performances are evaluated and compared to traditional artificial neural network. The comparison results indicate that the proposed modelling framework can increase the accuracy by 4%, and 26. 08%, reduce the uncertainty by 4% and at least 22. 15%, and improve the stability by 24. 09% and by 46. 34%, for one year and multi-year predictions, respectively. This framework offers great potential for accurate, reliable and stable predictions for engineering asset conditions by leveraging the complementary strengths of numerical simulations and data-driven models.

Details DOI

NeurIPS Conference 2025 Conference Paper

Self-supervised Blending Structural Context of Visual Molecules for Robust Drug Interaction Prediction

Tengfei Ma
Kun Chen
Yongsheng Zang
Yujie Chen
Xuanbai Ren
Bosheng Song
Hongxin Xiang
Yiping Liu

Identifying drug-drug interactions (DDIs) is critical for ensuring drug safety and advancing drug development, a topic that has garnered significant research interest. While existing methods have made considerable progress, approaches relying solely on known DDIs face a key challenge when applied to drugs with limited data: insufficient exploration of the space of unlabeled pairwise drugs. To address these issues, we innovatively introduce S$^2$VM, a Self-supervised Visual pretraining framework for pair-wise Molecules, to fully fuse structural representations and explore the space of drug pairs for DDI prediction. S$^2$VM incorporates the explicit structure and correlations of visual molecules, such as the positional relationships and connectivity between functional substructures. Specifically, we blend the visual fragments of drug pairs into a unified input for joint encoding and then recover molecule-specific visual information for each drug individually. This approach integrates fine-grained structural representations from unlabeled drug pair data. By using visual fragments as anchors, S$^2$VM effectively captures the spatial information of local molecular components within visual molecules, resulting in more comprehensive embeddings of drug pairs. Experimental results show that S$^2$VM achieves state-of-the-art performance on widely used benchmarks, with Macro-F1 score improvements of 4. 21% and 3. 31%, respectively. Further extensive results and theoretical analysis demonstrate the effectiveness of S$^2$VM for both few-shot and novel drugs.

PDF Details

TMLR Journal 2024 Journal Article

A Multilinear Least-Squares Formulation for Sparse Tensor Canonical Correlation Analysis

Jun Yu
Zhaoming Kong
Kun Chen
Xin Zhang
Yong Chen
Lifang He

Tensor data are becoming important recently in various applications, e.g., image and video recognition, which pose new challenges for data modeling and analysis approaches, such as high-order relations of large complexity, varying data scale and gross noise. In this paper, we consider the problem of sparse canonical correlation analysis for arbitrary tensor data. Although several methods have been proposed for this task, there are still limitations hindering its practical applications. To this end, we present a general Sparse Tensor Canonical Correlation Analysis (gSTCCA) method from a multilinear least-squares perspective. Specifically, we formulate the problem as a constrained multilinear least-squares problem with tensor-structured sparsity regularization based on CANDECOMP/PARAFAC (CP) decomposition. Then we present a divide-and-conquer deflation approach to tackle the problem by successive rank-one tensor estimation of the residual tensors, where the overall model is broken up into a set of unconstrained linear least-squares problems that can be efficiently solved. Through extensive experiments conducted on five different datasets for recognition tasks, we demonstrate that the proposed method achieves promising performance compared to the SOTA vector- and tensor-based canonical correlation analysis methods in terms of classification accuracy, model sparsity, and robustness to missing and noisy data. The code is publicly available at https://github.com/junfish/gSTCCA.

PDF Details

AIIM Journal 2024 Journal Article

EEG spatial inter-channel connectivity analysis: A GCN-based dual stream approach to distinguish mental fatigue status

Kun Chen
Shulong Chai
Tianli Xie
Quan Liu
Li Ma

Mental fatigue is defined as a decline in the ability and efficiency of mental activities. A lot of research suggests that the transition from alertness to fatigue is accompanied by alterations in correlation patterns among various brain regions. However, conventional methods for detecting mental fatigue seldom emphases inter-channel connectivity in the spatial domain. To fill this gap, this paper explores the spatial inter-channel connectivity in alertness and fatigue, employing spectral graph convolutional networks (GCN) for mental fatigue detection. We utilized Pearson correlation coefficients (PCC) to establish temporal connections and magnitude-squared coherence (MSC) for spectral connections. Topological features of the brain network were then analysed. To enhance the learning of spatial inter-channel connectivity, a dual-graph strategy transforms edge features into node features, serving as inputs to the spectral GCN. By simultaneously learning PCC and MSC features, the model results indicate significant differences in some brain network characteristics between alert and fatigue states. It confirms that the synchronicity of brain operations differs in the alert state compared to mental fatigue, and indicates that fatigue states can influence correlation patterns among different brain regions. Our approach is evaluated on a self-designed experimental dataset containing 7 subjects, demonstrating a classification accuracy of 89. 59 % in group-level experiments and 95. 24 % at the subject level. Additionally, on the public dataset SEED-VIG containing 23 subjects, our method achieves an accuracy of 86. 58 %. In summary, this paper proposes a neural network approach based on a dynamic functional connectivity network. The network integrates both temporal and spectral connections with the goal of simultaneously learning spatial inter-channel connectivity in time and frequency domains. This effectively accomplishes fatigue state detection, highlighting that fatigue significantly influences correlations among different brain regions.

Details DOI

NeurIPS Conference 2024 Conference Paper

FNP: Fourier Neural Processes for Arbitrary-Resolution Data Assimilation

Kun Chen
Peng Ye
Hao Chen
Kang Chen
Tao Han
Wanli Ouyang
Tao Chen
Lei Bai

Data assimilation is a vital component in modern global medium-range weather forecasting systems to obtain the best estimation of the atmospheric state by combining the short-term forecast and observations. Recently, AI-based data assimilation approaches have attracted increasing attention for their significant advantages over traditional techniques in terms of computational consumption. However, existing AI-based data assimilation methods can only handle observations with a specific resolution, lacking the compatibility and generalization ability to assimilate observations with other resolutions. Considering that complex real-world observations often have different resolutions, we propose the Fourier Neural Processes (FNP) for arbitrary-resolution data assimilation in this paper. Leveraging the efficiency of the designed modules and flexible structure of neural processes, FNP achieves state-of-the-art results in assimilating observations with varying resolutions, and also exhibits increasing advantages over the counterparts as the resolution and the amount of observations increase. Moreover, our FNP trained on a fixed resolution can directly handle the assimilation of observations with out-of-distribution resolutions and the observational information reconstruction task without additional fine-tuning, demonstrating its excellent generalization ability across data resolutions as well as across tasks. Code is available at https: //github. com/OpenEarthLab/FNP.

PDF Details DOI

JMLR Journal 2022 Journal Article

Fast Stagewise Sparse Factor Regression

Kun Chen
Ruipeng Dong
Wanwan Xu
Zemin Zheng

Sparse factorization of a large matrix is fundamental in modern statistical learning. In particular, the sparse singular value decomposition has been utilized in many multivariate regression methods. The appeal of this factorization is owing to its power in discovering a highly-interpretable latent association network. However, many existing methods are either ad hoc without a general performance guarantee, or are computationally intensive. We formulate the statistical problem as a sparse factor regression and tackle it with a two-stage “deflation + stagewise learning” approach. In the first stage, we consider both sequential and parallel approaches for simplifying the task into a set of co-sparse unit-rank estimation (CURE) problems, and establish the statistical underpinnings of these commonly-adopted and yet poorly understood deflation methods. In the second stage, we innovate a contended stagewise learning technique, consisting of a sequence of simple incremental updates, to efficiently trace out the whole solution paths of CURE. Our algorithm achieves a much lower computational complexity than alternating convex search, and it enables a flexible and principled tradeoff between statistical accuracy and computational efficiency. Our work is among the first to enable stagewise learning for non-convex problems, and the idea can be applicable in many multi-convex problems. Extensive simulation studies and an application in genetics demonstrate the effectiveness and scalability of our approach. [abs] [ pdf ][ bib ] &copy JMLR 2022. ( edit, beta )

PDF Details

ICML Conference 2022 Conference Paper

On Non-local Convergence Analysis of Deep Linear Networks

Kun Chen
Dachao Lin
Zhihua Zhang 0004

In this paper, we study the non-local convergence properties of deep linear networks. Specifically, under the quadratic loss, we consider optimizing deep linear networks in which there is at least a layer with only one neuron. We describe the convergent point of trajectories with an arbitrary balanced starting point under gradient flow, including the paths which converge to one of the saddle points. We also show specific convergence rates of trajectories that converge to the global minimizers by stages. We conclude that the rates vary from polynomial to linear. As far as we know, our results are the first to give a non-local analysis of deep linear neural networks with arbitrary balanced initialization, rather than the lazy training regime which has dominated the literature on neural networks or the restricted benign initialization.

Details

ICML Conference 2021 Conference Paper

Communication-Efficient Distributed SVD via Local Power Iterations

Xiang Li 0050
Shusen Wang
Kun Chen
Zhihua Zhang 0004

We study distributed computing of the truncated singular value decomposition (SVD). We develop an algorithm that we call \texttt{LocalPower} for improving communication efficiency. Specifically, we uniformly partition the dataset among $m$ nodes and alternate between multiple (precisely $p$) local power iterations and one global aggregation. In the aggregation, we propose to weight each local eigenvector matrix with orthogonal Procrustes transformation (OPT). As a practical surrogate of OPT, sign-fixing, which uses a diagonal matrix with $\pm 1$ entries as weights, has better computation complexity and stability in experiments. We theoretically show that under certain assumptions \texttt{LocalPower} lowers the required number of communications by a factor of $p$ to reach a constant accuracy. We also show that the strategy of periodically decaying $p$ helps obtain high-precision solutions. We conduct experiments to demonstrate the effectiveness of \texttt{LocalPower}.

Details

JBHI Journal 2020 Journal Article

Dehaze of Cataractous Retinal Images Using an Unpaired Generative Adversarial Network

Yuhao Luo
Kun Chen
Lei Liu
Jicheng Liu
Jianbo Mao
Genjie Ke
Mingzhai Sun

Cataracts are the leading cause of visual impairment worldwide. Examination of the retina through cataracts using a fundus camera is challenging and error-prone due to degraded image quality. We sought to develop an algorithm to dehaze such images to support diagnosis by either ophthalmologists or computer-aided diagnosis systems. Based on the generative adversarial network (GAN) concept, we designed two neural networks: CataractSimGAN and CataractDehazeNet. CataractSimGAN was intended for the synthesis of cataract-like images through unpaired clear retinal images and cataract images. CataractDehazeNet was trained using pairs of synthesized cataract-like images and the corresponding clear images through supervised learning. With two networks trained independently, the number of hyper-parameters was reduced, leading to better performance. We collected 400 retinal images without cataracts and 400 hazy images from cataract patients as the training dataset. Fifty cataract images and the corresponding clear images from the same patients after surgery comprised the test dataset. The clear images after surgery were used for reference to evaluate the performance of our method. CataractDehazeNet was able to enhance the degraded image from cataract patients substantially and to visualize blood vessels and the optic disc, while actively suppressing the artifacts common in application of similar methods. Thus, we developed an algorithm to improve the quality of the retinal images acquired from cataract patients. We achieved high structure similarity and fidelity between processed images and images from the same patients after cataract surgery.

Details DOI

IJCAI Conference 2018 Conference Paper

Beyond the Click-Through Rate: Web Link Selection with Multi-level Feedback

Kun Chen
Kechao Cai
Longbo Huang
John C. S. Lui

The web link selection problem is to select a small subset of web links from a large web link pool, and to place the selected links on a web page that can only accommodate a limited number of links, e. g. , advertisements, recommendations, or news feeds. Despite the long concerned click-through rate which reflects the attractiveness of the link itself, revenue can only be obtained from user actions after clicks, e. g. , purchasing after being directed to the product pages by recommendation links. Thus, web links have an intrinsic multi-level feedback structure. With this observation, we consider the context-free web link selection problem, where the objective is to maximize revenue while ensuring that the attractiveness is no less than a preset threshold. The key challenge of the problem is that each link's multi-level feedbacks are stochastic, and unobservable unless the link is selected. We model this problem with a constrained stochastic multi-armed bandit formulation, and design an efficient link selection algorithm, called Constrained Upper Confidence Bound algorithm (Con-UCB). We prove O(sqrt(T ln(T))) bounds on both regret and violation of the attractiveness constraint. We also conduct extensive experiments on three real-world datasets, and show that Con-UCB outperforms state-of-the-art context-free bandit algorithms concerning the multi-level feedback structure.

PDF Details

NeurIPS Conference 2018 Conference Paper

Boosted Sparse and Low-Rank Tensor Regression

Lifang He
Kun Chen
Wanwan Xu
Jiayu Zhou
Fei Wang

We propose a sparse and low-rank tensor regression model to relate a univariate outcome to a feature tensor, in which each unit-rank tensor from the CP decomposition of the coefficient tensor is assumed to be sparse. This structure is both parsimonious and highly interpretable, as it implies that the outcome is related to the features through a few distinct pathways, each of which may only involve subsets of feature dimensions. We take a divide-and-conquer strategy to simplify the task into a set of sparse unit-rank tensor regression problems. To make the computation efficient and scalable, for the unit-rank tensor regression, we propose a stagewise estimation procedure to efficiently trace out its entire solution path. We show that as the step size goes to zero, the stagewise solution paths converge exactly to those of the corresponding regularized regression. The superior performance of our approach is demonstrated on various real-world and synthetic examples.

PDF Details

YNIMG Journal 2010 Journal Article

A neurocomputational method for fully automated 3D dendritic spine detection and segmentation of medium-sized spiny neurons

Yong Zhang
Kun Chen
Matthew Baron
Merilee A. Teylan
Yong Kim
Zhihuan Song
Paul Greengard
Stephen T.C. Wong

Acquisition and quantitative analysis of high resolution images of dendritic spines are challenging tasks but are necessary for the study of animal models of neurological and psychiatric diseases. Currently available methods for automated dendritic spine detection are for the most part customized for 2D image slices, not volumetric 3D images. In this work, a fully automated method is proposed to detect and segment dendritic spines from 3D confocal microscopy images of medium-sized spiny neurons (MSNs). MSNs constitute a major neuronal population in striatum, and abnormalities in their function are associated with several neurological and psychiatric diseases. Such automated detection is critical for the development of new 3D neuronal assays which can be used for the screening of drugs and the studies of their therapeutic effects. The proposed method utilizes a generalized gradient vector flow (GGVF) with a new smoothing constraint and then detects feature points near the central regions of dendrites and spines. Then, the central regions are refined and separated based on eigen-analysis and multiple shape measurements. Finally, the spines are segmented in 3D space using the fast marching algorithm, taking the detected central regions of spines as initial points. The proposed method is compared with three popular existing methods for centerline extraction and also with manual results for dendritic spine detection in 3D space. The experimental results and comparisons show that the proposed method is able to automatically and accurately detect, segment, and quantitate dendritic spines in 3D images of MSNs.

Details DOI