Rapid Parameter Estimation for Extreme Mass Ratio Inspirals Using Machine Learning

Bo Liang,Hong Guo,Tianyu Zhao,He wang,Herik Evangelinelis,Yuxiang Xu,Chang liu,Manjia Liang,Xiaotong Wei,Yong Yuan,Peng Xu,Minghui Du,Wei-Liang Qian,Ziren Luo

2024-09-12

Abstract:Extreme-mass-ratio inspiral (EMRI) signals pose significant challenges in gravitational wave (GW) astronomy owing to their low-frequency nature and highly complex waveforms, which occupy a high-dimensional parameter space with numerous variables. Given their extended inspiral timescales and low signal-to-noise ratios, EMRI signals warrant prolonged observation periods. Parameter estimation becomes particularly challenging due to non-local parameter degeneracies, arising from multiple local maxima, as well as flat regions and ridges inherent in the likelihood function. These factors lead to exceptionally high time complexity for parameter analysis while employing traditional matched filtering and random sampling methods. To address these challenges, the present study applies machine learning to Bayesian posterior estimation of EMRI signals, leveraging the recently developed flow matching technique based on ODE neural networks. Our approach demonstrates computational efficiency several orders of magnitude faster than the traditional Markov Chain Monte Carlo (MCMC) methods, while preserving the unbiasedness of parameter estimation. We show that machine learning technology has the potential to efficiently handle the vast parameter space, involving up to seventeen parameters, associated with EMRI signals. Furthermore, to our knowledge, this is the first instance of applying machine learning, specifically the Continuous Normalizing Flows (CNFs), to EMRI signal analysis. Our findings highlight the promising potential of machine learning in EMRI waveform analysis, offering new perspectives for the advancement of space-based GW detection and GW astronomy.

Computational Physics,Instrumentation and Methods for Astrophysics,Artificial Intelligence

What problem does this paper attempt to address?

The paper aims to address the problem of parameter estimation for Extreme Mass Ratio Inspirals (EMRIs) signals. EMRI signals are of significant importance in gravitational wave astronomy, but due to their low-frequency characteristics and complex waveform structures, parameter estimation faces numerous challenges. Specifically, the parameter space of EMRI signals is high-dimensional, with many variables, and there exists non-local parameter degeneracy, making traditional methods such as Markov Chain Monte Carlo (MCMC) methods inefficient and inaccurate for parameter estimation.

To solve these problems, the study proposes a machine learning-based method utilizing the recently developed Flow Matching technique based on Ordinary Differential Equation Neural Networks (ODE Neural Networks). This method not only improves computational efficiency but also effectively avoids getting trapped in local optima. Through experimental validation, this method demonstrates higher efficiency in handling EMRI signals compared to traditional MCMC methods while maintaining unbiased parameter estimation. Additionally, the study shows that machine learning techniques can effectively manage the large parameter space even with up to 17 parameters. Overall, the study provides an innovative solution for the analysis of future real EMRI signals.