首页|Reports from Guizhou Normal University Advance Knowledge in Intelligent Systems (Molecular Subgraph Representation Learning Based On Spatial Structure Transform er)
Reports from Guizhou Normal University Advance Knowledge in Intelligent Systems (Molecular Subgraph Representation Learning Based On Spatial Structure Transform er)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ma chine Learning-Intelligent Systems. According to news reporting originating fr om Guizhou, People's Republic of China, by NewsRx correspondents, research state d, "In the field of molecular biology, graph representation learning is crucial for molecular structure analysis. However, challenges arise in recognising funct ional groups and distinguishing isomers due to a lack of spatial structure infor mation." Funders for this research include The Science and Technology Foundation of Guizh ou Province, Science and Technology Foundation of Guizhou Province, Guizhou Prov incial Key Technology R D Program. Our news editors obtained a quote from the research from Guizhou Normal Universi ty, "To address these problems, we design a novel graph representation learning method based on a spatial structure information extraction Transformer (SSET). T he SSET model comprises the Edge Feature Fusion Subgraph Spatial Structure Extra ctor (ETSE) module and the Positional Information Encoding Graph Transformer (PE GT) module. The ETSE module extracts spatial structural information by fusing ed ge features and generating the most-value subgraph (Mv-subgraph). The PEGT modul e encodes positional information based on the graph transformer, addressing the indistinguishability problem among nodes with identical features. In addition, t he SSET model alleviates the burden of high computational complexity by using su bgraph."
GuizhouPeople's Republic of ChinaAsi aIntelligent SystemsMachine LearningGuizhou Normal University
NETL
NSTL
