Rapid parameter estimation for merging massive black hole binaries using continuous normalizing flows
文献类型：期刊论文
| 作者 | Liang, Bo; Du MH(杜明辉); Wang, He; Xu YX(许宇翔); Liu, Chang ; Wei XT(魏晓通); Xu P(徐鹏); Qiang, Lie; Luo ZR(罗子人)
|
| 刊名 | MACHINE LEARNING-SCIENCE AND TECHNOLOGY
 |
| 出版日期 | 2024-12 |
| 卷号 | 5期号:4页码:45040 |
| 关键词 | gravitational wave massive black hole binaries continuous normalizing flows flow matching |
| DOI | 10.1088/2632-2153/ad8da9 |
| 英文摘要 | Detecting the coalescences of massive black hole binaries (MBHBs) is one of the primary targets for space-based gravitational wave observatories such as laser interferometer space antenna, Taiji, and Tianqin. The fast and accurate parameter estimation of merging MBHBs is of great significance for the global fitting of all resolvable sources, as well as the astrophysical interpretation of gravitational wave signals. However, such analyses usually entail significant computational costs. To address these challenges, inspired by the latest progress in generative models, we explore the application of continuous normalizing flows (CNFs) on the parameter estimation of MBHBs. Specifically, we employ linear interpolation and trig interpolation methods to construct transport paths for training CNFs. Additionally, we creatively introduce a parameter transformation method based on the symmetry in the detector's response function. This transformation is integrated within CNFs, allowing us to train the model using a simplified dataset, and then perform parameter estimation on more general data, hence also acting as a crucial factor in improving the training speed. In conclusion, for the first time, within a comprehensive and reasonable parameter range, we have achieved a complete and unbiased 11-dimensional rapid inference for MBHBs in the presence of astrophysical confusion noise using CNFs. In the experiments based on simulated data, our model produces posterior distributions comparable to those obtained by nested sampling. |
| 分类号 | 二类/Q1 |
| WOS研究方向 | Computer Science ; Science & Technology - Other Topics |
| 语种 | 英语 |
| WOS记录号 | WOS:001354502000001 |
| 资助机构 | International Partnership Program of the Chinese Academy of Sciences ; National Key Research and Development Program of China {2021YFC2201901, 2021YFC2203004, 2020YFC2200100, 2021YFC2201903]${025GJHZ2023106GC] |
| 其他责任者 | Du MH ; Wang H |
| 源URL | [http://dspace.imech.ac.cn/handle/311007/97161]  |
| 专题 | 力学研究所_国家微重力实验室 |
| 作者单位 | 1.【Liu, Chang & Qiang, Li-e】 Chinese Acad Sci, Natl Space Sci Ctr, Beijing 100190, Peoples R China 2.【Wang, He & Luo, Ziren】 Univ Chinese Acad Sci UCAS, Int Ctr Theoret Phys Asia Pacific ICTP AP, Beijing 100049, Peoples R China 3.【Xu, Peng】 Lanzhou Univ, Lanzhou Ctr Theoret Phys, Lanzhou 730000, Peoples R China 4.【Liang, Bo & Wang, He & Xu, Yuxiang & Xu, Peng & Luo, Ziren】 Univ Chinese Acad Sci UCAS, Taiji Lab Gravitat Wave Univ Beijing Hangzhou, Beijing 100049, Peoples R China 5.【Liang, Bo & Xu, Yuxiang】 Chinese Acad Sci, Shanghai Inst Opt & Fine Mech, Shanghai 201800, Peoples R China 6.【Liang, Bo & Xu, Yuxiang & Xu, Peng & Luo, Ziren】 UCAS, Hangzhou Inst Adv Study, Key Lab Gravitat Wave Precis Measurement Zhejiang, Hangzhou 310024, Peoples R China 7.【Liang, Bo & Du, Minghui & Xu, Yuxiang & Wei, Xiaotong & Xu, Peng & Luo, Ziren】 Chinese Acad Sci, Inst Mech, Ctr Gravitat Wave Expt, Natl Micrograv Lab, Beijing 100190, Peoples R China |
| 推荐引用方式 GB/T 7714 | Liang, Bo,Du MH,Wang, He,et al. Rapid parameter estimation for merging massive black hole binaries using continuous normalizing flows[J]. MACHINE LEARNING-SCIENCE AND TECHNOLOGY,2024,5(4):45040. |
| APA | Liang, Bo.,杜明辉.,Wang, He.,许宇翔.,Liu, Chang.,...&罗子人.(2024).Rapid parameter estimation for merging massive black hole binaries using continuous normalizing flows.MACHINE LEARNING-SCIENCE AND TECHNOLOGY,5(4),45040. |
| MLA | Liang, Bo,et al."Rapid parameter estimation for merging massive black hole binaries using continuous normalizing flows".MACHINE LEARNING-SCIENCE AND TECHNOLOGY 5.4(2024):45040. |
入库方式： OAI收割
来源：力学研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。