arXiv subjects

Xiuling Wang

Publications and source records attributed to Xiuling Wang.

12 recordsLinked to original sources

arXiv2026

Communication-efficient Federated Graph Classification via Generative Diffusion Modeling

Graph Neural Networks (GNNs) unlock new ways of learning from graph-structured data, proving highly effective in capturing complex relationships and patterns. Federated GNNs (FGNNs) have emerged as a prominent distributed learning paradigm for training GNNs over decentralized data. However, FGNNs face two significant challenges: high communication overhead from multiple rounds of parameter exchanges and non-IID data characteristics across clients. To address these issues, we introduce CeFGC, a novel FGNN paradigm that facilitates efficient GNN training over non-IID data by limiting communication between the server and clients to three rounds only. The core idea of CeFGC is to leverage generative diffusion models to minimize direct client-server communication. Each client trains a generative diffusion model that captures its local graph distribution and shares this model with the server, which then redistributes it back to all clients. Using these generative models, clients generate synthetic graphs combined with their local graphs to train local GNN models. Finally, clients upload their model weights to the server for aggregation into a global GNN model. We theoretically analyze the I/O complexity of communication volume to show that CeFGC reduces to a constant of three communication rounds only. Extensive experiments on several real graph datasets demonstrate the effectiveness and efficiency of CeFGC against state-of-the-art competitors, reflecting our superior performance on non-IID graphs by aligning local and global model objectives and enriching the training set with diverse graphs.

cs.LG↗

arXiv2026

Inference Attacks Against Graph Generative Diffusion Models

Graph generative diffusion models have recently emerged as a powerful paradigm for generating complex graph structures, effectively capturing intricate dependencies and relationships within graph data. However, the privacy risks associated with these models remain largely unexplored. In this paper, we investigate information leakage in such models through three types of black-box inference attacks. First, we design a graph reconstruction attack, which can reconstruct graphs structurally similar to those training graphs from the generated graphs. Second, we propose a property inference attack to infer the properties of the training graphs, such as the average graph density and the distribution of densities, from the generated graphs. Third, we develop two membership inference attacks to determine whether a given graph is present in the training set. Extensive experiments on three different types of graph generative diffusion models and six real-world graphs demonstrate the effectiveness of these attacks, significantly outperforming the baseline approaches. Finally, we propose two defense mechanisms that mitigate these inference attacks and achieve a better trade-off between defense strength and target model utility than existing methods. Our code is available at https://zenodo.org/records/17946102.

cs.LG↗

arXiv2026

Feature-Aware One-Shot Federated Learning via Hierarchical Token Sequences

One-shot federated learning (OSFL) reduces the communication cost and privacy risks of iterative federated learning by constructing a global model with a single round of communication. However, most existing methods struggle to achieve robust performance on real-world domains such as medical imaging, or are inefficient when handling non-IID (Independent and Identically Distributed) data. To address these limitations, we introduce FALCON, a framework that enhances the effectiveness of OSFL over non-IID image data. The core idea of FALCON is to leverage the feature-aware hierarchical token sequences generation and knowledge distillation into OSFL. First, each client leverages a pretrained visual encoder with hierarchical scale encoding to compress images into hierarchical token sequences, which capture multi-scale semantics. Second, a multi-scale autoregressive transformer generator is used to model the distribution of these token sequences and generate the synthetic sequences. Third, clients upload the synthetic sequences along with the local classifier trained on the real token sequences to the server. Finally, the server incorporates knowledge distillation into global training to reduce reliance on precise distribution modeling. Experiments on medical and natural image datasets validate the effectiveness of FALCON in diverse non-IID scenarios, outperforming the best OSFL baselines by 9.58% in average accuracy.

cs.LG↗

arXiv2024

MPCPA: Multi-Center Privacy Computing with Predictions Aggregation based on Denoising Diffusion Probabilistic Model

Privacy-preserving computing is crucial for multi-center machine learning in many applications such as healthcare and finance. In this paper a Multi-center Privacy Computing framework with Predictions Aggregation (MPCPA) based on denoising diffusion probabilistic model (DDPM) is proposed, in which conditional diffusion model training, DDPM data generation, a classifier, and strategy of prediction aggregation are included. Compared to federated learning, this framework necessitates fewer communications and leverages high-quality generated data to support robust privacy computing. Experimental validation across multiple datasets demonstrates that the proposed framework outperforms classic federated learning and approaches the performance of centralized learning with original data. Moreover, our approach demonstrates robust security, effectively addressing challenges such as image memorization and membership inference attacks. Our experiments underscore the efficacy of the proposed framework in the realm of privacy computing, with the code set to be released soon.

cs.DC↗

arXiv2023

Equipping Federated Graph Neural Networks with Structure-aware Group Fairness

Graph Neural Networks (GNNs) have been widely used for various types of graph data processing and analytical tasks in different domains. Training GNNs over centralized graph data can be infeasible due to privacy concerns and regulatory restrictions. Thus, federated learning (FL) becomes a trending solution to address this challenge in a distributed learning paradigm. However, as GNNs may inherit historical bias from training data and lead to discriminatory predictions, the bias of local models can be easily propagated to the global model in distributed settings. This poses a new challenge in mitigating bias in federated GNNs. To address this challenge, we propose $\text{F}^2$GNN, a Fair Federated Graph Neural Network, that enhances group fairness of federated GNNs. As bias can be sourced from both data and learning algorithms, $\text{F}^2$GNN aims to mitigate both types of bias under federated settings. First, we provide theoretical insights on the connection between data bias in a training graph and statistical fairness metrics of the trained GNN models. Based on the theoretical analysis, we design $\text{F}^2$GNN which contains two key components: a fairness-aware local model update scheme that enhances group fairness of the local models on the client side, and a fairness-weighted global model update scheme that takes both data bias and fairness metrics of local models into consideration in the aggregation process. We evaluate $\text{F}^2$GNN empirically versus a number of baseline methods, and demonstrate that $\text{F}^2$GNN outperforms these baselines in terms of both fairness and model accuracy.

cs.LG↗

arXiv2022

Group Property Inference Attacks Against Graph Neural Networks

With the fast adoption of machine learning (ML) techniques, sharing of ML models is becoming popular. However, ML models are vulnerable to privacy attacks that leak information about the training data. In this work, we focus on a particular type of privacy attacks named property inference attack (PIA) which infers the sensitive properties of the training data through the access to the target ML model. In particular, we consider Graph Neural Networks (GNNs) as the target model, and distribution of particular groups of nodes and links in the training graph as the target property. While the existing work has investigated PIAs that target at graph-level properties, no prior works have studied the inference of node and link properties at group level yet. In this work, we perform the first systematic study of group property inference attacks (GPIA) against GNNs. First, we consider a taxonomy of threat models under both black-box and white-box settings with various types of adversary knowledge, and design six different attacks for these settings. We evaluate the effectiveness of these attacks through extensive experiments on three representative GNN models and three real-world graphs. Our results demonstrate the effectiveness of these attacks whose accuracy outperforms the baseline approaches. Second, we analyze the underlying factors that contribute to GPIA's success, and show that the target model trained on the graphs with or without the target property represents some dissimilarity in model parameters and/or model outputs, which enables the adversary to infer the existence of the property. Further, we design a set of defense mechanisms against the GPIA attacks, and demonstrate that these mechanisms can reduce attack accuracy effectively with small loss on GNN model accuracy.

cs.LG↗

arXiv2016

Convex PBW-type Lyndon Bases and Restricted Two-Parameter Quantum Group of Type $F_4$

We determine convex PBW-type Lyndon bases for two-parameter quantum groups $U_{r,s}(F_4)$ with detailed commutation relations. We construct a finite-dimensional Hopf algebra $\mathfrak u_{r,s}(F_4)$, as a quotient of $U_{r,s}(F_4)$ by a Hopf ideal generated by certain central elements, which is pointed, and of a Drinfel'd double structure under a certain condition. All of Hopf isomorphisms of $\mathfrak u_{r,s}(F_4)$ are determined which are important for seeking the possible new pointed objects in low order with $(\ell, 210)\ne 1$. Finally, necessary and sufficient conditions for $\mathfrak u_{r,s}(F_4)$ to be a ribbon Hopf algebra are singled out by describing the left and right integrals.

math.QA↗

arXiv2012

Modular quantizations of Lie algebras of Cartan type $H$ via Drinfeld Twists

We construct explicit Drinfel'd twists for the generalized Cartan type $H$ Lie algebras in characteristic $0$ and obtain the corresponding quantizations and their integral forms. Via making modular reductions including modulo $p$ reduction and modulo $p$-restrictedness reduction, and base changes, we derive certain modular quantizations of the restricted universal enveloping algebra $\mathbf u(\mathbf{H}(2n;\underline{1}))$ in characteristic $p$. They are new non-pointed Hopf algebras of truncated $p$-polynomial noncommutative and noncocommutative deformation of prime-power dimension $p^{p^{2n}-1}$, which contain the well-known Radford algebras as Hopf subalgebras. As a by-product, we also get some Jordanian quantizations for $\mathfrak {sp}_{2n}$.

math.QA↗

arXiv2009

Convex PBW-type Lyndon Basis and Restricted Two-parameter Quantum Group of Type G_2

We construct finite-dimensional pointed Hopf algebras \mathfrak u_{r,s}(G_2) (i.e. restricted 2-parameter quantum groups) from the 2-parameter quantum group U_{r,s}(G_2) defined in \cite{HS}, which turn out to be of Drinfel'd doubles, where a crucial point is to give a detailed combinatorial construction of the convex PBW-type Lyndon basis for type G_2 in 2-parameter quantum version. After furnishing possible commutation relations among quantum root vectors, we show that the restricted quantum groups are ribbon Hopf algebras under certain conditions through determining their left and right integrals. Besides these, we determine all of the Hopf algebra isomorphisms of u_{r,s}(G_2) in terms of the description of the sets of its left (right) skew-primitive elements.

math.QA↗

arXiv2009

Quantizations of generalized Cartan type $S$ Lie algebras and of the special algebra $\mathbf{S}(n;\underline{1})$ in the modular case

The generalized Cartan type $\mathbf{S}$ Lie algebras in char 0 with the Lie bialgebra structures involved are quantized, where the Drinfel'd twist we used is proved to be a variation of the Jordanian twist. As the passage from char 0 to char p, their quantization integral forms are given. By the modular reduction and base changes, we obtain certain quantizations of the restricted universal enveloping algebra $\mathbf u(\mathbf{S}(n;\underline{1}))$ (for the Cartan type simple modular restricted Lie algebra $\mathbf{S}(n;\underline{1})$ of $\mathbf{S}$ type). They are new Hopf algebras of truncated $p$-polynomial noncommutative and noncocommutative deformation of dimension $p^{1+(n-1)(p^n-1)}$, which contain the well-known Radford algebra (\cite{DR}) as a Hopf subalgebra. As a by-product, we also get some Jordanian quantizations for $\mathfrak {sl}_n$, which are induced from those horizontal quantizations of $\mathbf S(n;\underline1)$.

math.QA↗

arXiv2008

Convex PBW-Type Lyndon Bases and Restricted Two-parameter Quantum Groups of Type B

We construct convex PBW-type Lyndon bases for two-parameter quantum groups U_{r,s}({so}_{2n+1}) with detailed commutation relations. It turns out that under a certain condition, the restricted two-parameter quantum group u_{r,s}({so}_{2n+1}) (r, s are roots of unity) is of Drinfeld double. All of Hopf isomorphisms of u_{r,s}({so}_{2n+1}), as well as u_{r,s}({sl}_n) are determined. Finally, necessary and sufficient conditions for u_{r,s}({so}_{2n+1}) to be a ribbon Hopf algebra are singled out by describing the left and right integrals.

math.QA↗

arXiv2007

Quantizations of generalized-Witt algebra and of Jacobson-Witt algebra in the modular case

We quantize the generalized-Witt algebra in characteristic 0 with its Lie bialgebra structures discovered by Song-Su (\cite{GY}). Via a modulo p reduction and a modulo "p-restrictedness" reduction process, we get 2^n{-}1 families of truncated p-polynomial noncocommutative deformations of the restricted universal enveloping algebra of the Jacobson-Witt algebra \mathbf{W}(n;\underline{1}) (for the Cartan type simple modular restricted Lie algebra of W type). They are new families of noncommutative and noncocommutative Hopf algebras of dimension p^{1+np^n} in characteristic p. Our results generalize a work of Grunspan (J. Algebra 280 (2004), 145--161, \cite{CG}) in rank n=1 case in characteristic 0. In the modular case, the argument for a refined version follows from the modular reduction approach (different from \cite{CG}) with some techniques from the modular Lie algebra theory.

math.QA↗