A5101547511- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Graph Theory and Algorithms
- Recommender Systems and Techniques
- Topic Modeling
- Opinion Dynamics and Social Influence
- Domain Adaptation and Few-Shot Learning
- Machine Learning in Materials Science
- Multimodal Machine Learning Applications
- Computational Drug Discovery Methods
- Service-Oriented Architecture and Web Services
- Resource-Constrained Project Scheduling
- Advanced Manufacturing and Logistics Optimization
- Innovative Microfluidic and Catalytic Techniques Innovation
- Electrocatalysts for Energy Conversion
- Network Security and Intrusion Detection
- Advanced Computational Techniques and Applications
- Mesoporous Materials and Catalysis
- Crystallization and Solubility Studies
- Scheduling and Optimization Algorithms
- Seed and Plant Biochemistry
- Heavy Metals in Plants
- Machine Learning and ELM
- Machine Learning in Healthcare
- Cocoa and Sweet Potato Agronomy
Wuhan University
2014-2024
City University of Hong Kong, Shenzhen Research Institute
2020-2024
Hubei University Of Economics
2024
Tianjin University
2022
National University of Singapore
2022
Anhui Polytechnic University
2021
Nanjing University
2015
Tropical Crops Genetic Resources Institute
2014
Chinese Academy of Tropical Agricultural Sciences
2014
China Three Gorges University
2012
Graph neural networks have emerged as a leading architecture for many graph-level tasks, such graph classification and generation. As an essential component of the architecture, pooling is indispensable obtaining holistic representation whole graph. Although great variety methods been proposed in this promising fast-developing research field, to best our knowledge, little effort has made systematically summarize these works. To set stage development future works, paper, we attempt fill gap...
Recently, social networks have witnessed a massive surge in popularity. A key issue network research is evolution analysis, which assumes that all the autonomous nodes follow uniform mechanisms. However, different should mechanisms to generate edges. This proposed as underlying idea ensure nodes' diversity this paper. Our approach involves identifying micro-level node generates edges by introducing existing link prediction methods from perspectives of nodes. We also propose edge generation...
Node classification is a fundamental graph-based task that aims to predict the classes of unlabeled nodes, for which Graph Neural Networks (GNNs) are state-of-the-art methods. Current GNNs assume nodes in training set contribute equally during training. However, quality varies greatly, and performance could be harmed by two types low-quality nodes: (1) inter-class situated near class boundaries lack typical characteristics their corresponding classes. Because data-driven approaches, on these...
Graph Transformers (GTs) have proved their advantage in graph-level tasks. However, existing GTs still perform unsatisfactorily on the node classification task due to 1) overwhelming unrelated information obtained from a vast number of irrelevant distant nodes and 2) quadratic complexity regarding via fully connected attention mechanism. In this paper, we present Gapformer, method for that deeply incorporates Transformer with Pooling. More specifically, Gapformer coarsens large-scale graph...
Online proton range/dose verification based on measurements of proton-induced positron emitters is a promising strategy for quality assurance in therapy. Because the nonlinear correlation between dose distribution and activity emitters, we aim to establish their relationship using recurrent neural network models (LSTM, BiLSTM, GRU, BiGRU Seq2seq). Simulations were carried out with spot-scanning system Geant4-10.3 toolkit CT-based patient phantom. The 1D distributions radiation obtained....
Yuren Mao, Zekai Wang, Weiwei Liu, Xuemin Lin, Wenbin Hu. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
Abstract Side scan sonar measurement platform, affected by underwater environment and its own motion precision, inevitably has posture disturbance, which greatly affects accuracy of geomorphic image formation. It is difficult to sensitively accurately capture these disturbances relying on auxiliary navigation devices. In this paper, we propose a method invert information the platform using matching relation between strip images. The inversion algorithm key link in mosaic frame side sonar,...
How can we discover and estimate major events in complex social networks? Event detection evaluation networks provide an effective solution, which has become the critical basis for many real applications, such as crisis management decision making. However, existing methods ignore difference of evolution fluctuations nodes. In order to further improve accuracy event detection, this paper proposes method based on node (NodeED). It contains a similarity index algorithm (SimJudge) microevolution...
Oxygen Electrocatalysis In article number 2202215, Xuerong Zheng, Xiaopeng Han, and co-workers report a class of Janus homologous heterostructures (metal alloys/sulfides) via an ultrafast high-temperature shock strategy. Benefiting from the synergistic effect metallic sites in interfaces, exhibit enhanced oxygen catalytic performance impressive durability as efficient air cathode electrocatalyst for rechargeable zinc-air batteries.
Mass customization-producing mode had higher demands to product designing courses. Product collaborative design system based on mass customization was designed in this paper, an architecture J2EE and engineering database proposed. The data of the were divided into global local forms, while further classified as management data. Management carried out using JSP+JavaBean technology. An Applet would communicate with I-DEAS under CORBA protocol Web style is a good resolution project enterprise...
In order to enhance the ability of multi-agent system deal with conflicts complex system, negotiation mechanism was presented based on n-person nonzero-sum game theory in this paper, which provided course descriptions negotiation, process and thread, ft composing arithmetic aimed at alliance base importing bilateral self-value. It adopted Nash concept verify stability negotiating arithmetic, demonstrated that under meaning balance.
Signed network embedding (SNE) has received considerable attention in recent years. A mainstream idea of SNE is to learn node representations by estimating the ratio sampling densities. Though achieving promising performance, these methods based on density estimation are limited issues confusing sample, expected error, and fixed priori. To alleviate above-mentioned issues, this paper, we propose a novel dual-branch (DDRE) architecture for SNE. Specifically, DDRE 1) consists network, dealing...