Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning

doi:10.3390/e23040403

	Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning
	Zhang, Xun 1; Yang, Lanyan 1; Zhang, Bin 1; Liu, Ying 1; Jiang, Dong 2; Qin, Xiaohai 1; Hao, Mengmeng 2
刊名	ENTROPY
	2021-04-01
卷号	23 期号:4 页码:17
关键词	graph analysis graph neural network semi-supervised learning neighborhood aggregation
DOI	10.3390/e23040403
通讯作者	Liu, Ying(liu_ying@th.btbu.edu.cn) ; Jiang, Dong(jiangd@igsnrr.ac.cn)
英文摘要	The problem of extracting meaningful data through graph analysis spans a range of different fields, such as social networks, knowledge graphs, citation networks, the World Wide Web, and so on. As increasingly structured data become available, the importance of being able to effectively mine and learn from such data continues to grow. In this paper, we propose the multi-scale aggregation graph neural network based on feature similarity (MAGN), a novel graph neural network defined in the vertex domain. Our model provides a simple and general semi-supervised learning method for graph-structured data, in which only a very small part of the data is labeled as the training set. We first construct a similarity matrix by calculating the similarity of original features between all adjacent node pairs, and then generate a set of feature extractors utilizing the similarity matrix to perform multi-scale feature propagation on graphs. The output of multi-scale feature propagation is finally aggregated by using the mean-pooling operation. Our method aims to improve the model representation ability via multi-scale neighborhood aggregation based on feature similarity. Extensive experimental evaluation on various open benchmarks shows the competitive performance of our method compared to a variety of popular architectures.
资助项目	National Key Research and Development Program of China[2020YFB1806500] ; Support Project of High-level Teachers in Beijing Municipal Universities in the Period of 13th Five-year Plan[CITTCD201904037] ; R&D Program of Beijing Municipal Education Commission[KM202010011012]
WOS研究方向	Physics
语种	英语
出版者	MDPI
WOS记录号	WOS:000642997100001
资助机构	National Key Research and Development Program of China ; Support Project of High-level Teachers in Beijing Municipal Universities in the Period of 13th Five-year Plan ; R&D Program of Beijing Municipal Education Commission
内容类型	期刊论文
源URL	[http://ir.igsnrr.ac.cn/handle/311030/161608]
专题	中国科学院地理科学与资源研究所
通讯作者	Liu, Ying; Jiang, Dong
作者单位	1.Beijing Technol & Business Univ, Sch Comp Sci & Engn, Beijing Key Lab Big Data Technol Food Safety, Beijing 100048, Peoples R China 2.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Key Lab Resources Utilizat & Environm Remediat, Beijing 100101, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Xun,Yang, Lanyan,Zhang, Bin,et al. Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning[J]. ENTROPY,2021,23(4):17.
APA	Zhang, Xun.,Yang, Lanyan.,Zhang, Bin.,Liu, Ying.,Jiang, Dong.,...&Hao, Mengmeng.(2021).Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning.ENTROPY,23(4),17.
MLA	Zhang, Xun,et al."Multi-Scale Aggregation Graph Neural Networks Based on Feature Similarity for Semi-Supervised Learning".ENTROPY 23.4(2021):17.