用近似梯度拟合随机格模型

Jan Schering, Sander Keemink, Johannes Textor
{"title":"用近似梯度拟合随机格模型","authors":"Jan Schering, Sander Keemink, Johannes Textor","doi":"arxiv-2310.08305","DOIUrl":null,"url":null,"abstract":"Stochastic lattice models (sLMs) are computational tools for simulating\nspatiotemporal dynamics in physics, computational biology, chemistry, ecology,\nand other fields. Despite their widespread use, it is challenging to fit sLMs\nto data, as their likelihood function is commonly intractable and the models\nnon-differentiable. The adjacent field of agent-based modelling (ABM), faced\nwith similar challenges, has recently introduced an approach to approximate\ngradients in network-controlled ABMs via reparameterization tricks. This\napproach enables efficient gradient-based optimization with automatic\ndifferentiation (AD), which allows for a directed local search of suitable\nparameters rather than estimation via black-box sampling. In this study, we\ninvestigate the feasibility of using similar reparameterization tricks to fit\nsLMs through backpropagation of approximate gradients. We consider four common\nscenarios: fitting to single-state transitions, fitting to trajectories,\ninference of lattice states, and identification of stable lattice\nconfigurations. We demonstrate that all tasks can be solved by AD using four\nexample sLMs from sociology, biophysics, image processing, and physical\nchemistry. Our results show that AD via approximate gradients is a promising\nmethod to fit sLMs to data for a wide variety of models and tasks.","PeriodicalId":501231,"journal":{"name":"arXiv - PHYS - Cellular Automata and Lattice Gases","volume":"34 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2023-10-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Fitting Stochastic Lattice Models Using Approximate Gradients\",\"authors\":\"Jan Schering, Sander Keemink, Johannes Textor\",\"doi\":\"arxiv-2310.08305\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Stochastic lattice models (sLMs) are computational tools for simulating\\nspatiotemporal dynamics in physics, computational biology, chemistry, ecology,\\nand other fields. Despite their widespread use, it is challenging to fit sLMs\\nto data, as their likelihood function is commonly intractable and the models\\nnon-differentiable. The adjacent field of agent-based modelling (ABM), faced\\nwith similar challenges, has recently introduced an approach to approximate\\ngradients in network-controlled ABMs via reparameterization tricks. This\\napproach enables efficient gradient-based optimization with automatic\\ndifferentiation (AD), which allows for a directed local search of suitable\\nparameters rather than estimation via black-box sampling. In this study, we\\ninvestigate the feasibility of using similar reparameterization tricks to fit\\nsLMs through backpropagation of approximate gradients. We consider four common\\nscenarios: fitting to single-state transitions, fitting to trajectories,\\ninference of lattice states, and identification of stable lattice\\nconfigurations. We demonstrate that all tasks can be solved by AD using four\\nexample sLMs from sociology, biophysics, image processing, and physical\\nchemistry. Our results show that AD via approximate gradients is a promising\\nmethod to fit sLMs to data for a wide variety of models and tasks.\",\"PeriodicalId\":501231,\"journal\":{\"name\":\"arXiv - PHYS - Cellular Automata and Lattice Gases\",\"volume\":\"34 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-10-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Cellular Automata and Lattice Gases\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2310.08305\",\"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 - PHYS - Cellular Automata and Lattice Gases","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2310.08305","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

随机晶格模型(sLMs)是模拟物理、计算生物学、化学、生态学和其他领域的时空动力学的计算工具。尽管它们被广泛使用,但由于它们的似然函数通常难以处理且模型不可微,因此将它们拟合到数据中是具有挑战性的。邻近的基于智能体的建模(ABM)领域面临着类似的挑战,最近引入了一种通过重新参数化技巧在网络控制的ABM中近似梯度的方法。这种方法实现了有效的基于梯度的自动微分优化(AD),它允许对合适参数进行定向局部搜索,而不是通过黑盒采样进行估计。在这项研究中,我们研究了通过近似梯度的反向传播使用类似的重新参数化技巧来拟合slm的可行性。我们考虑了四种常见的场景:拟合到单态转换,拟合到轨迹,推断晶格状态,以及识别稳定的晶格构型。我们使用社会学、生物物理学、图像处理和物理化学的四个示例slm来证明所有任务都可以通过AD解决。我们的结果表明,通过近似梯度的AD是一种很有前途的方法,可以将slm拟合到各种模型和任务的数据中。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Fitting Stochastic Lattice Models Using Approximate Gradients
Stochastic lattice models (sLMs) are computational tools for simulating spatiotemporal dynamics in physics, computational biology, chemistry, ecology, and other fields. Despite their widespread use, it is challenging to fit sLMs to data, as their likelihood function is commonly intractable and the models non-differentiable. The adjacent field of agent-based modelling (ABM), faced with similar challenges, has recently introduced an approach to approximate gradients in network-controlled ABMs via reparameterization tricks. This approach enables efficient gradient-based optimization with automatic differentiation (AD), which allows for a directed local search of suitable parameters rather than estimation via black-box sampling. In this study, we investigate the feasibility of using similar reparameterization tricks to fit sLMs through backpropagation of approximate gradients. We consider four common scenarios: fitting to single-state transitions, fitting to trajectories, inference of lattice states, and identification of stable lattice configurations. We demonstrate that all tasks can be solved by AD using four example sLMs from sociology, biophysics, image processing, and physical chemistry. Our results show that AD via approximate gradients is a promising method to fit sLMs to data for a wide variety of models and tasks.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信