Searcharxiv⌕ Search

arXiv subjects

Yao Ma

Publications and source records attributed to Yao Ma.

At least 109 records · Page 6Linked to original sources
arXiv2024

A Survey on Graph Condensation

Analytics on large-scale graphs have posed significant challenges to computational efficiency and resource requirements. Recently, Graph condensation (GC) has emerged as a solution to address challenges arising from the escalating volume of graph data. The motivation of GC is to reduce the scale of large graphs to smaller ones while preserving essential information for downstream tasks. For a better understanding of GC and to distinguish it from other related topics, we present a formal definition of GC and establish a taxonomy that systematically categorizes existing methods into three types based on its objective, and classify the formulations to generate the condensed graphs into two categories as modifying the original graphs or synthetic completely new ones. Moreover, our survey includes a comprehensive analysis of datasets and evaluation metrics in this field. Finally, we conclude by addressing challenges and limitations, outlining future directions, and offering concise guidelines to inspire future research in this field.

cs.LG↗
arXiv2023

Tetraquark bound states in constituent quark models: benchmark test calculations

We investigate the tetraquark bound states that are manifestly exotic using three distinct few-body methods: Gaussian Expansion Method (GEM), Resonating Group Method (RGM), and Diffusion Monte Carlo (DMC). We refer to manifestly exotic states that do not involve a mixture with the conventional mesons through the creation and annihilation of $n\bar{n}$, where $n=u, d$. Our calculations are conducted with two types of quark models: the pure constituent quark model featuring one-gluon-exchange interactions and confinement interactions, and the chiral constituent quark model, supplemented by extra one-boson-exchange interactions. This study represents a comprehensive benchmark test of various few-body methods and quark models. Our findings reveal the superiority of GEM over RGM and DMC methods based on present implements for the tetraquark bound states. Additionally, we observe a tendency for the chiral quark model to overestimate the binding energies. We systematically explore the fully, triply, doubly, and singly heavy tetraquark states with $J^P=0^+,1^+,2^+$, encompassing over 150 states in total. We successfully identify several bound states, including $[cc\bar{n}\bar{n}]_{J^{P}=1^{+}}^{I=0}$, $[bb\bar{n}\bar{n}]_{J^{P}=1^{+}}^{I=0}$, $[bc\bar{n}\bar{n}]_{J^{P}=0^{+},1^{+},2^{+}}^{I=0}$, $[bs\bar{n}\bar{n}]_{J^{P}=0^{+},1^{+}}^{I=0}$, $[cs\bar{n}\bar{n}]_{J^{P}=0^{+}}^{I=0}$, and $[bb\bar{n}\bar{s}]_{J^{P}=1^{+}}$, all found to be bound states below the dimeson thresholds.

hep-ph↗
arXiv2023

Evaluating Graph Neural Networks for Link Prediction: Current Pitfalls and New Benchmarking

Link prediction attempts to predict whether an unseen edge exists based on only a portion of edges of a graph. A flurry of methods have been introduced in recent years that attempt to make use of graph neural networks (GNNs) for this task. Furthermore, new and diverse datasets have also been created to better evaluate the effectiveness of these new models. However, multiple pitfalls currently exist that hinder our ability to properly evaluate these new methods. These pitfalls mainly include: (1) Lower than actual performance on multiple baselines, (2) A lack of a unified data split and evaluation metric on some datasets, and (3) An unrealistic evaluation setting that uses easy negative samples. To overcome these challenges, we first conduct a fair comparison across prominent methods and datasets, utilizing the same dataset and hyperparameter search settings. We then create a more practical evaluation setting based on a Heuristic Related Sampling Technique (HeaRT), which samples hard negative samples via multiple heuristics. The new evaluation setting helps promote new challenges and opportunities in link prediction by aligning the evaluation with real-world situations. Our implementation and data are available at https://github.com/Juanhui28/HeaRT

cs.LG↗
arXiv2023

Graph Enhanced BERT for Query Understanding

Query understanding plays a key role in exploring users' search intents and facilitating users to locate their most desired information. However, it is inherently challenging since it needs to capture semantic information from short and ambiguous queries and often requires massive task-specific labeled data. In recent years, pre-trained language models (PLMs) have advanced various natural language processing tasks because they can extract general semantic information from large-scale corpora. Therefore, there are unprecedented opportunities to adopt PLMs for query understanding. However, there is a gap between the goal of query understanding and existing pre-training strategies -- the goal of query understanding is to boost search performance while existing strategies rarely consider this goal. Thus, directly applying them to query understanding is sub-optimal. On the other hand, search logs contain user clicks between queries and urls that provide rich users' search behavioral information on queries beyond their content. Therefore, in this paper, we aim to fill this gap by exploring search logs. In particular, to incorporate search logs into pre-training, we first construct a query graph where nodes are queries and two queries are connected if they lead to clicks on the same urls. Then we propose a novel graph-enhanced pre-training framework, GE-BERT, which can leverage both query content and the query graph. In other words, GE-BERT can capture both the semantic information and the users' search behavioral information of queries. Extensive experiments on various query understanding tasks have demonstrated the effectiveness of the proposed framework.

cs.IR↗
arXiv2023

Soft photon theorem in QCD with massless quarks

Working to all orders in dimensionally-regularized QCD, we study the radiation of a photon whose energy is much lower than that of external partons, but much larger than the masses of some quarks. We argue that the conventional soft photon theorem receives corrections at leading power in the photon energy, associated with soft virtual loops of massless fermions. These additive corrections give an overall factor times the non-radiative amplitude that is infrared finite and real to all orders in $α_s$. Based on recent calculations of the three-loop soft gluon current, we identify the lowest-order three-loop correction.

hep-ph↗
arXiv2023

Distance-Based Propagation for Efficient Knowledge Graph Reasoning

Knowledge graph completion (KGC) aims to predict unseen edges in knowledge graphs (KGs), resulting in the discovery of new facts. A new class of methods have been proposed to tackle this problem by aggregating path information. These methods have shown tremendous ability in the task of KGC. However they are plagued by efficiency issues. Though there are a few recent attempts to address this through learnable path pruning, they often sacrifice the performance to gain efficiency. In this work, we identify two intrinsic limitations of these methods that affect the efficiency and representation quality. To address the limitations, we introduce a new method, TAGNet, which is able to efficiently propagate information. This is achieved by only aggregating paths in a fixed window for each source-target pair. We demonstrate that the complexity of TAGNet is independent of the number of layers. Extensive experiments demonstrate that TAGNet can cut down on the number of propagated messages by as much as 90% while achieving competitive performance on multiple KG datasets. The code is available at https://github.com/HarryShomer/TAGNet.

cs.LG↗
arXiv2023

Demystifying Structural Disparity in Graph Neural Networks: Can One Size Fit All?

Recent studies on Graph Neural Networks(GNNs) provide both empirical and theoretical evidence supporting their effectiveness in capturing structural patterns on both homophilic and certain heterophilic graphs. Notably, most real-world homophilic and heterophilic graphs are comprised of a mixture of nodes in both homophilic and heterophilic structural patterns, exhibiting a structural disparity. However, the analysis of GNN performance with respect to nodes exhibiting different structural patterns, e.g., homophilic nodes in heterophilic graphs, remains rather limited. In the present study, we provide evidence that Graph Neural Networks(GNNs) on node classification typically perform admirably on homophilic nodes within homophilic graphs and heterophilic nodes within heterophilic graphs while struggling on the opposite node set, exhibiting a performance disparity. We theoretically and empirically identify effects of GNNs on testing nodes exhibiting distinct structural patterns. We then propose a rigorous, non-i.i.d PAC-Bayesian generalization bound for GNNs, revealing reasons for the performance disparity, namely the aggregated feature distance and homophily ratio difference between training and testing nodes. Furthermore, we demonstrate the practical implications of our new findings via (1) elucidating the effectiveness of deeper GNNs; and (2) revealing an over-looked distribution shift factor on graph out-of-distribution problem and proposing a new scenario accordingly.

cs.LG↗
arXiv2023

Single-soft emissions for amplitudes with two colored particles at three loops

We compute the three-loop correction to the universal single-soft emission current for the case of scattering amplitudes with two additional color-charged partons. We present results valid for QCD and $\mathcal{N}=4$ super-symmetric Yang-Mills theory. To achieve our results we develop a new integrand expansion technique for scattering amplitudes in the presence of soft emissions. Furthermore, we obtain contributions from single final-state parton matrix elements to the Higgs boson and Drell-Yan production cross section at next-to-next-to-next-to-next-to leading order (N$^4$LO) in perturbative QCD in the threshold limit.

hep-ph↗
arXiv2023

Multi-Instance Adversarial Attack on GNN-Based Malicious Domain Detection

Malicious domain detection (MDD) is an open security challenge that aims to detect if an Internet domain is associated with cyber-attacks. Among many approaches to this problem, graph neural networks (GNNs) are deemed highly effective. GNN-based MDD uses DNS logs to represent Internet domains as nodes in a maliciousness graph (DMG) and trains a GNN to infer their maliciousness by leveraging identified malicious domains. Since this method relies on accessible DNS logs to construct DMGs, it exposes a vulnerability for adversaries to manipulate their domain nodes' features and connections within DMGs. Existing research mainly concentrates on threat models that manipulate individual attacker nodes. However, adversaries commonly generate multiple domains to achieve their goals economically and avoid detection. Their objective is to evade discovery across as many domains as feasible. In this work, we call the attack that manipulates several nodes in the DMG concurrently a multi-instance evasion attack. We present theoretical and empirical evidence that the existing single-instance evasion techniques for are inadequate to launch multi-instance evasion attacks against GNN-based MDDs. Therefore, we introduce MintA, an inference-time multi-instance adversarial attack on GNN-based MDDs. MintA enhances node and neighborhood evasiveness through optimized perturbations and operates successfully with only black-box access to the target model, eliminating the need for knowledge about the model's specifics or non-adversary nodes. We formulate an optimization challenge for MintA, achieving an approximate solution. Evaluating MintA on a leading GNN-based MDD technique with real-world data showcases an attack success rate exceeding 80%. These findings act as a warning for security experts, underscoring GNN-based MDDs' susceptibility to practical attacks that can undermine their effectiveness and benefits.

cs.CR↗
arXiv2023

Towards Fair Graph Neural Networks via Graph Counterfactual

Graph neural networks have shown great ability in representation (GNNs) learning on graphs, facilitating various tasks. Despite their great performance in modeling graphs, recent works show that GNNs tend to inherit and amplify the bias from training data, causing concerns of the adoption of GNNs in high-stake scenarios. Hence, many efforts have been taken for fairness-aware GNNs. However, most existing fair GNNs learn fair node representations by adopting statistical fairness notions, which may fail to alleviate bias in the presence of statistical anomalies. Motivated by causal theory, there are several attempts utilizing graph counterfactual fairness to mitigate root causes of unfairness. However, these methods suffer from non-realistic counterfactuals obtained by perturbation or generation. In this paper, we take a causal view on fair graph learning problem. Guided by the casual analysis, we propose a novel framework CAF, which can select counterfactuals from training data to avoid non-realistic counterfactuals and adopt selected counterfactuals to learn fair node representations for node classification task. Extensive experiments on synthetic and real-world datasets show the effectiveness of CAF. Our code is available at https://github.com/TimeLovercc/CAF-GNN.

cs.LG↗
arXiv2023

Is Homophily a Necessity for Graph Neural Networks?

Graph neural networks (GNNs) have shown great prowess in learning representations suitable for numerous graph-based machine learning tasks. When applied to semi-supervised node classification, GNNs are widely believed to work well due to the homophily assumption ("like attracts like"), and fail to generalize to heterophilous graphs where dissimilar nodes connect. Recent works design new architectures to overcome such heterophily-related limitations, citing poor baseline performance and new architecture improvements on a few heterophilous graph benchmark datasets as evidence for this notion. In our experiments, we empirically find that standard graph convolutional networks (GCNs) can actually achieve better performance than such carefully designed methods on some commonly used heterophilous graphs. This motivates us to reconsider whether homophily is truly necessary for good GNN performance. We find that this claim is not quite true, and in fact, GCNs can achieve strong performance on heterophilous graphs under certain conditions. Our work carefully characterizes these conditions, and provides supporting theoretical understanding and empirical observations. Finally, we examine existing heterophilous graphs benchmarks and reconcile how the GCN (under)performs on them based on this understanding.

cs.LG↗
arXiv2023

Are Message Passing Neural Networks Really Helpful for Knowledge Graph Completion?

Knowledge graphs (KGs) facilitate a wide variety of applications. Despite great efforts in creation and maintenance, even the largest KGs are far from complete. Hence, KG completion (KGC) has become one of the most crucial tasks for KG research. Recently, considerable literature in this space has centered around the use of Message Passing (Graph) Neural Networks (MPNNs), to learn powerful embeddings. The success of these methods is naturally attributed to the use of MPNNs over simpler multi-layer perceptron (MLP) models, given their additional message passing (MP) component. In this work, we find that surprisingly, simple MLP models are able to achieve comparable performance to MPNNs, suggesting that MP may not be as crucial as previously believed. With further exploration, we show careful scoring function and loss function design has a much stronger influence on KGC model performance. This suggests a conflation of scoring function design, loss function design, and MP in prior work, with promising insights regarding the scalability of state-of-the-art KGC methods today, as well as careful attention to more suitable MP designs for KGC tasks tomorrow. Our codes are publicly available at: https://github.com/Juanhui28/Are_MPNNs_helpful.

cs.AI↗
arXiv2023

Quantum Lock: A Provable Quantum Communication Advantage

Physical unclonable functions(PUFs) provide a unique fingerprint to a physical entity by exploiting the inherent physical randomness. Gao et al. discussed the vulnerability of most current-day PUFs to sophisticated machine learning-based attacks. We address this problem by integrating classical PUFs and existing quantum communication technology. Specifically, this paper proposes a generic design of provably secure PUFs, called hybrid locked PUFs(HLPUFs), providing a practical solution for securing classical PUFs. An HLPUF uses a classical PUF(CPUF), and encodes the output into non-orthogonal quantum states to hide the outcomes of the underlying CPUF from any adversary. Here we introduce a quantum lock to protect the HLPUFs from any general adversaries. The indistinguishability property of the non-orthogonal quantum states, together with the quantum lockdown technique prevents the adversary from accessing the outcome of the CPUFs. Moreover, we show that by exploiting non-classical properties of quantum states, the HLPUF allows the server to reuse the challenge-response pairs for further client authentication. This result provides an efficient solution for running PUF-based client authentication for an extended period while maintaining a small-sized challenge-response pairs database on the server side. Later, we support our theoretical contributions by instantiating the HLPUFs design using accessible real-world CPUFs. We use the optimal classical machine-learning attacks to forge both the CPUFs and HLPUFs, and we certify the security gap in our numerical simulation for construction which is ready for implementation.

quant-ph↗
arXiv2023

Ground state baryons in the flux-tube three-body confinement model using Diffusion Monte Carlo

We make a systematical diffusion Monte Carlo (DMC) calculation for all ground state baryons in two confinement scenarios, the pairwise confinement and the three-body flux-tube confinement. With the baryons as an example, we illustrate a feasible procedure to investigate the few-quark states with possible few-body confinement mechanisms, which can be extended to the multiquark states easily. For each baryon, we extract the mass, mean-square radius, charge radius, and the quark distributions. We use the Jackknife resampling method to estimate the statistical uncertainties of masses to be less than 1 MeV. To determine the baryon charge radii, we include the constituent quark size effect, which is fixed by the experimental and lattice QCD results. Our results show that both two-body and three-body confinement mechanisms can give a good description of the experimental data if the parameters are chosen properly. In the flux-tube confinement, introducing different tension parameters for the baryons and mesons are necessary, specifically, $σ_Y= 0.9204 σ_{Q\bar{Q}}$. The lesson from the calculation of the nucleon mass with the DMC method is that the improper pre-assignment of the channels may prevent us from obtaining the real ground state. With this experience, we obtain the real ground state (the $η_c η_c$ threshold with the di-meson configuration) of the $cc\bar{c}\bar{c}$ system with $J^{PC}=0^{++}$ starting from the diquark-antidiquark spin-color channels alone, which is hard to achieve in the variational method and was not obtained in the previous DMC calculations.

hep-ph↗
arXiv2022

The on-shell expansion: from Landau equations to the Newton polytope

We study the application of the method of regions to Feynman integrals with massless propagators contributing to off-shell Green's functions in Minkowski spacetime (with non-exceptional momenta) around vanishing external masses, $p_i^2\to 0$. This on-shell expansion allows us to identify all infrared-sensitive regions at any power, in terms of infrared subgraphs in which a subset of the propagators become collinear to external lightlike momenta and others become soft. We show that each such region can be viewed as a solution to the Landau equations, or equivalently, as a facet in the Newton polytope constructed from the Symanzik graph polynomials. This identification allows us to study the properties of the graph polynomials associated with infrared regions, as well as to construct a graph-finding algorithm for the on-shell expansion, which identifies all regions using exclusively graph-theoretical conditions. We also use the results to investigate the analytic structure of integrals associated with regions in which every connected soft subgraph connects to just two jets. For such regions we prove that multiple on-shell expansions commute. This applies in particular to all regions in Sudakov form-factor diagrams as well as in any planar diagram.

hep-th↗
arXiv2022

Heterogeneous Randomized Response for Differential Privacy in Graph Neural Networks

Graph neural networks (GNNs) are susceptible to privacy inference attacks (PIAs), given their ability to learn joint representation from features and edges among nodes in graph data. To prevent privacy leakages in GNNs, we propose a novel heterogeneous randomized response (HeteroRR) mechanism to protect nodes' features and edges against PIAs under differential privacy (DP) guarantees without an undue cost of data and model utility in training GNNs. Our idea is to balance the importance and sensitivity of nodes' features and edges in redistributing the privacy budgets since some features and edges are more sensitive or important to the model utility than others. As a result, we derive significantly better randomization probabilities and tighter error bounds at both levels of nodes' features and edges departing from existing approaches, thus enabling us to maintain high data utility for training GNNs. An extensive theoretical and empirical analysis using benchmark datasets shows that HeteroRR significantly outperforms various baselines in terms of model utility under rigorous privacy protection for both nodes' features and edges. That enables us to defend PIAs in DP-preserving GNNs effectively.

cs.LG↗
arXiv2022

Learning Representations for Hyper-Relational Knowledge Graphs

Knowledge graphs (KGs) have gained prominence for their ability to learn representations for uni-relational facts. Recently, research has focused on modeling hyper-relational facts, which move beyond the restriction of uni-relational facts and allow us to represent more complex and real-world information. However, existing approaches for learning representations on hyper-relational KGs majorly focus on enhancing the communication from qualifiers to base triples while overlooking the flow of information from base triple to qualifiers. This can lead to suboptimal qualifier representations, especially when a large amount of qualifiers are presented. It motivates us to design a framework that utilizes multiple aggregators to learn representations for hyper-relational facts: one from the perspective of the base triple and the other one from the perspective of the qualifiers. Experiments demonstrate the effectiveness of our framework for hyper-relational knowledge graph completion across multiple datasets. Furthermore, we conduct an ablation study that validates the importance of the various components in our framework. The code to reproduce our results can be found at \url{https://github.com/HarryShomer/QUAD}.

cs.LG↗
arXiv2022

Cohomology and deformations of module homorphisms

In this paper, we mainly focus on formal deformation theory of module homomorphisms. We first introduce the cohomology of module homomorphisms and study formal one-parameter deformation. We obtain some properties about obstructions. Then we give some examples of deformations of modules and module homomorphisms.

math.RA↗