{"title":"STAResNet:用于求解麦克斯韦 PDE 的时空代数网络","authors":"Alberto Pepe, Sven Buchholz, Joan Lasenby","doi":"arxiv-2408.13619","DOIUrl":null,"url":null,"abstract":"We introduce STAResNet, a ResNet architecture in Spacetime Algebra (STA) to\nsolve Maxwell's partial differential equations (PDEs). Recently, networks in\nGeometric Algebra (GA) have been demonstrated to be an asset for truly\ngeometric machine learning. In \\cite{brandstetter2022clifford}, GA networks\nhave been employed for the first time to solve partial differential equations\n(PDEs), demonstrating an increased accuracy over real-valued networks. In this\nwork we solve Maxwell's PDEs both in GA and STA employing the same ResNet\narchitecture and dataset, to discuss the impact that the choice of the right\nalgebra has on the accuracy of GA networks. Our study on STAResNet shows how\nthe correct geometric embedding in Clifford Networks gives a mean square error\n(MSE), between ground truth and estimated fields, up to 2.6 times lower than\nthan obtained with a standard Clifford ResNet with 6 times fewer trainable\nparameters. STAREsNet demonstrates consistently lower MSE and higher\ncorrelation regardless of scenario. The scenarios tested are: sampling period\nof the dataset; presence of obstacles with either seen or unseen\nconfigurations; the number of channels in the ResNet architecture; the number\nof rollout steps; whether the field is in 2D or 3D space. This demonstrates how\nchoosing the right algebra in Clifford networks is a crucial factor for more\ncompact, accurate, descriptive and better generalising pipelines.","PeriodicalId":501309,"journal":{"name":"arXiv - CS - Computational Engineering, Finance, and Science","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"STAResNet: a Network in Spacetime Algebra to solve Maxwell's PDEs\",\"authors\":\"Alberto Pepe, Sven Buchholz, Joan Lasenby\",\"doi\":\"arxiv-2408.13619\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We introduce STAResNet, a ResNet architecture in Spacetime Algebra (STA) to\\nsolve Maxwell's partial differential equations (PDEs). Recently, networks in\\nGeometric Algebra (GA) have been demonstrated to be an asset for truly\\ngeometric machine learning. In \\\\cite{brandstetter2022clifford}, GA networks\\nhave been employed for the first time to solve partial differential equations\\n(PDEs), demonstrating an increased accuracy over real-valued networks. In this\\nwork we solve Maxwell's PDEs both in GA and STA employing the same ResNet\\narchitecture and dataset, to discuss the impact that the choice of the right\\nalgebra has on the accuracy of GA networks. Our study on STAResNet shows how\\nthe correct geometric embedding in Clifford Networks gives a mean square error\\n(MSE), between ground truth and estimated fields, up to 2.6 times lower than\\nthan obtained with a standard Clifford ResNet with 6 times fewer trainable\\nparameters. STAREsNet demonstrates consistently lower MSE and higher\\ncorrelation regardless of scenario. The scenarios tested are: sampling period\\nof the dataset; presence of obstacles with either seen or unseen\\nconfigurations; the number of channels in the ResNet architecture; the number\\nof rollout steps; whether the field is in 2D or 3D space. This demonstrates how\\nchoosing the right algebra in Clifford networks is a crucial factor for more\\ncompact, accurate, descriptive and better generalising pipelines.\",\"PeriodicalId\":501309,\"journal\":{\"name\":\"arXiv - CS - Computational Engineering, Finance, and Science\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computational Engineering, Finance, and Science\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2408.13619\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computational Engineering, Finance, and Science","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2408.13619","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
我们介绍了 STAResNet,这是时空代数(STA)中的一种 ResNet 架构,用于求解麦克斯韦偏微分方程(PDE)。最近,几何代数(GA)中的网络已被证明是真正几何机器学习的资产。在《cite{brandstetter2022clifford}》一书中,GA网络首次被用于求解偏微分方程(PDEs),与实值网络相比,其准确性得到了提高。在这项研究中,我们采用相同的 ResNet 架构和数据集,在 GA 和 STA 中求解了麦克斯韦 PDE,并讨论了右代数的选择对 GA 网络准确性的影响。我们对 STAResNet 的研究表明,在克利福德网络中进行正确的几何嵌入后,地面实况与估计场之间的均方误差(MSE)比可训练参数少 6 倍的标准克利福德 ResNet 低 2.6 倍。STAREsNet 在任何情况下都表现出较低的 MSE 和较高的相关性。测试的场景包括:数据集的采样周期;存在可见或不可见配置的障碍物;ResNet 架构中的通道数量;滚动步骤的数量;场地是在二维空间还是三维空间。这说明了在克利福德网络中选择正确的代数是如何成为更紧凑、更准确、更有描述性和更有概括性的管道的关键因素。
STAResNet: a Network in Spacetime Algebra to solve Maxwell's PDEs
We introduce STAResNet, a ResNet architecture in Spacetime Algebra (STA) to
solve Maxwell's partial differential equations (PDEs). Recently, networks in
Geometric Algebra (GA) have been demonstrated to be an asset for truly
geometric machine learning. In \cite{brandstetter2022clifford}, GA networks
have been employed for the first time to solve partial differential equations
(PDEs), demonstrating an increased accuracy over real-valued networks. In this
work we solve Maxwell's PDEs both in GA and STA employing the same ResNet
architecture and dataset, to discuss the impact that the choice of the right
algebra has on the accuracy of GA networks. Our study on STAResNet shows how
the correct geometric embedding in Clifford Networks gives a mean square error
(MSE), between ground truth and estimated fields, up to 2.6 times lower than
than obtained with a standard Clifford ResNet with 6 times fewer trainable
parameters. STAREsNet demonstrates consistently lower MSE and higher
correlation regardless of scenario. The scenarios tested are: sampling period
of the dataset; presence of obstacles with either seen or unseen
configurations; the number of channels in the ResNet architecture; the number
of rollout steps; whether the field is in 2D or 3D space. This demonstrates how
choosing the right algebra in Clifford networks is a crucial factor for more
compact, accurate, descriptive and better generalising pipelines.