arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1309.7709astro-ph.IMastro-ph.CO
收藏

基于马尔可夫链蒙特卡罗的高结构引力波信号后验估计

Efficient Estimation of Highly Structured Posteriors of Gravitational-Wave Signals with Markov-Chain Monte Carlo

Benjamin Farr, Vicky Kalogera, Erik Luijten

摘要:

我们介绍了一种新的马尔可夫链蒙特卡罗(MCMC)方法设计的高相关性和多模态后验的有效采样。平行回火虽然有效,但对于采样此类后验是一种昂贵的技术。我们的方法最大限度地减少了并行回火的使用,只在短时间内使用它来调整新的跳跃建议。对于复杂的后验,我们发现效率提高了~13倍。目前,对地基探测器测量的引力波信号的参数估计是通过贝叶斯推断和MCMC抽样方法来完成的。这些信号的后验通常是具有强非线性相关性的多模态,使得采样困难。随着我们进入高级检测器时代,灵敏度的提高和带宽的扩大将大大增加分析的计算成本,需要更有效的搜索算法来应对这些挑战。

英文摘要:

We introduce a new Markov-Chain Monte Carlo (MCMC) approach designed for efficient sampling of highly correlated and multimodal posteriors. Parallel tempering, though effective, is a costly technique for sampling such posteriors. Our approach minimizes the use of parallel tempering, only using it for a short time to tune a new jump proposal. For complex posteriors we find efficiency improvements up to a factor of ~13. The estimation of parameters of gravitational-wave signals measured by ground-based detectors is currently done through Bayesian inference with MCMC one of the leading sampling methods. Posteriors for these signals are typically multimodal with strong non-linear correlations, making sampling difficult. As we enter the advanced-detector era, improved sensitivities and wider bandwidths will drastically increase the computational cost of analyses, demanding more efficient search algorithms to meet these challenges.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑