A Survey of Graph Neural Networks and Their Industrial Applications DOI
Haoran Lu, Lei Wang, Xiaoliang Ma

et al.

Published: Jan. 1, 2024

Graph Neural Networks (GNNs) have emerged as a powerful tool for analyzing and modeling graph-structured data. In recent years, GNNs gained significant attention in various domains. This review paper aims to provide an overview of the state-of-the-art graph neural network techniques their industrial applications.First, we introduce fundamental concepts architectures GNNs, highlighting ability capture complex relationships dependencies We then delve into variants evolution graphs, including directed heterogeneous dynamic hypergraphs. Next, discuss interpretability GNN, GNN theory augmentation, expressivity, over-smoothing.Finally, specific use cases settings, finance, biology, knowledge recommendation systems, Internet Things (IoT), distillation. highlights immense potential solving real-world problems, while also addressing challenges opportunities further advancement this field.

Language: Английский

Machine Learning for Blockchain Data Analysis: Progress and Opportunities DOI
Poupak Azad, Cüneyt Gürcan Akçora, Arijit Khan

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: April 8, 2025

Blockchain technology has rapidly emerged to mainstream attention. At the same time, its publicly accessible, heterogeneous, massive-volume, and temporal data are reminiscent of complex dynamics encountered during last decade big data. Unlike any prior source, blockchain datasets encompass multiple layers interactions across real-world entities, e.g., human users, autonomous programs, smart contracts. Furthermore, blockchain's integration with cryptocurrencies introduced financial aspects unprecedented scale complexity, such as decentralized finance, stablecoins, non-fungible tokens, central bank digital currencies. These unique characteristics present opportunities challenges for machine learning on On one hand, we examine state-of-the-art solutions, applications, future directions associated leveraging analysis critical improving technology, e-crime detection trends prediction. other shed light pivotal role by providing vast tools that can catalyze growth evolving ecosystem. This paper is a comprehensive resource researchers, practitioners, policymakers, offering roadmap navigating this dynamic transformative field.

Language: Английский

Citations

0
Dance Video Action Recognition Algorithm Based on Improved Hypergraph Convolutional Networks DOI Creative Commons
Zhen Ni,

Yiyi Jiang

Systems and Soft Computing, Journal Year: 2025, Volume and Issue: unknown, P. 200247 - 200247

Published: April 1, 2025

Language: Английский

Citations

0
Generalization Performance of Hypergraph Neural Networks DOI
Yifan Wang, Gonzalo R. Arce, Guangmo Tong

et al.

Published: April 22, 2025

Language: Английский

Citations

0
Beyond Sequential Patterns: Rethinking Healthcare Predictions with Contextual Insights DOI
Chuang Zhao, Hui Tang, Hongke Zhao

et al.

ACM transactions on office information systems, Journal Year: 2025, Volume and Issue: unknown

Published: April 29, 2025

Healthcare predictions, such as readmission prediction, stand a cornerstone of societal well-being, exerting profound influence on individual health outcomes and communal vitality. Existing research primarily employs advanced graph neural networks sequential algorithms for patient modeling, with focus discerning the connections patterns inherent in Electronic Health Records (EHRs). However, heterogeneity entity interactions, locality EHR data, oversight target relevance hinder further improvements. To address these limitations, we introduce novel framework B eyond S equential P atterns (BSP), which facilitates precise healthcare predictions by incorporating tri-contextual information. Specifically, establish symptom-driven hypergraph network four semantic hyperedges tailored to intricacies scenario, ontology. This serves global context, tracking heterogeneous collaboration within across patients. Moreover, construct an extensive knowledge leveraging existing medical databases large language models. By sampling refining subgraphs local bolster associations entities from closed-set data open world. Finally, candidate explicit entity-relation loss. It enforces neighbor consistency between representation during optimization, thus accounting correlations among targets. Extensive experiments rigorous robustness analysis five tasks derived datasets underscore BSP’s superiority over leading baselines, improvements 11%, 3%, 3.5%, 2% tasks, demonstrating efficacy diverse contexts.

Language: Английский

Citations

0
A Survey of Graph Neural Networks and Their Industrial Applications DOI
Haoran Lu, Lei Wang, Xiaoliang Ma

et al.

Published: Jan. 1, 2024

Graph Neural Networks (GNNs) have emerged as a powerful tool for analyzing and modeling graph-structured data. In recent years, GNNs gained significant attention in various domains. This review paper aims to provide an overview of the state-of-the-art graph neural network techniques their industrial applications.First, we introduce fundamental concepts architectures GNNs, highlighting ability capture complex relationships dependencies We then delve into variants evolution graphs, including directed heterogeneous dynamic hypergraphs. Next, discuss interpretability GNN, GNN theory augmentation, expressivity, over-smoothing.Finally, specific use cases settings, finance, biology, knowledge recommendation systems, Internet Things (IoT), distillation. highlights immense potential solving real-world problems, while also addressing challenges opportunities further advancement this field.

Language: Английский

Citations

3