Please use this identifier to cite or link to this item:
http://oaps.umac.mo/handle/10692.1/359
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | CHOI, WAN IOI(蔡昀叡) | - |
dc.date.accessioned | 2024-07-16T08:54:08Z | - |
dc.date.available | 2024-07-16T08:54:08Z | - |
dc.date.issued | 2024 | - |
dc.identifier.citation | CHOI, W.I. (2024). Efficient Higher Order Self Attention Via Tensor Operations With Topology Enhanced Graph Neural Network For Molecular Graph Classification (Outstanding Academic Papers by Students (OAPS)). Retrieved from University of Macau, Outstanding Academic Papers by Students Repository. | en_US |
dc.identifier.uri | http://oaps.umac.mo/handle/10692.1/359 | - |
dc.description.abstract | We modify a recently proposed higher order generalization of self attention mechanism and reduce its time complexity using tensor operations. Then the modified higher order self attention is applied to a graph neural network with a novel structural encoding based on cell complex from algebraic topology. Experiments on molecular graph benchmarks show that our modified higher order self attention is more efficient than the original higher order self attention, and our proposed structural encoding improves the performance of the graph neural network. | en_US |
dc.language.iso | en | en_US |
dc.subject | self attention | en_US |
dc.subject | tensor decomposition | en_US |
dc.subject | molecular property prediction | en_US |
dc.subject | graph neural network | en_US |
dc.subject | cell complex | en_US |
dc.title | Efficient Higher Order Self Attention Via Tensor Operations With Topology Enhanced Graph Neural Network For Molecular Graph Classification | en_US |
dc.type | OAPS | en_US |
dc.contributor.department | Department of Mathematics | en_US |
dc.description.instructor | Prof. Kou Kit Ian | en_US |
dc.description.programme | Bachelor of Science in Mathematics (Mathematics and Applications Stream) | en_US |
Appears in Collections: | FST OAPS 2024 |
Files in This Item:
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.