引力波相位偏差的可识别性极限：多头神经网络的多分类
Identifiability Limits of Gravitational Wave Phase Deviations: Multiclass Classification with a Multihead Neural Network
浏览论文内容
中文总结 AI 辅助
本研究利用多头神经网络对引力波相位偏差进行分类与回归，发现识别极限主要源于相邻相位族的相似性，而非网络能力，为引力波检验提供了新见解。
中文摘要 AI 辅助
我们研究了使用具有分类和回归头的神经网络来识别引力波相位偏差的效果。分类头将广义相对论（GR）与六个参数化的后爱因斯坦相位族区分开来，这些相位族的指数 $b\in\{-7,-5,-3,-1,+1,+2\}$；回归头预测耦合幅度的对数 $\log_{10}|\beta|$。网络输入是一个响应函数，用于量化波形失配对波形变形的敏感度。我们使用具有先进LIGO设计谱的平稳高斯噪声，对262个源进行了测试，并使用第三次观测运行（O3）的实测汉福德谱对132个源进行了测试，同时也测试了注入到记录的汉福德应变中的信号。在两个高斯数据集上，网络以0.427和0.329的准确率清晰地在噪声之上识别出偏差的指数。为了解释这些结果，我们将网络与一个基于预测波形残差构建的近似分类器进行比较，该分类器针对六个相位族和相同的源。两个分类器经常混淆相同的族。通过检查不同相位修正留下的残差，我们将这些错误与潜在偏差之间的相似性联系起来。在选定的记录汉福德应变上，使用模拟或记录噪声训练的网络，其总体分类准确率接近模拟汉福德噪声上的结果。在先进LIGO设计数据集上测试的十三个神经网络变体中，循环网络在响亮偏差上具有最高的平均分类准确率，尽管所有变体仍低于近似分类器。网络几乎达到该分类器的准确率，表明对于已知的固有源参数，识别受到相邻相位族之间相似性的限制，而非网络本身的限制。
英文摘要
We study how well gravitational-wave phase deviations can be identified using a neural network with classification and regression heads. The classification head distinguishes general relativity (GR) from six parametrized post-Einsteinian phase families with exponents $b\in\{-7,-5,-3,-1,+1,+2\}$; the regression head predicts the logarithm of the coupling magnitude, $\log_{10}|β|$. The network input is a response function quantifying the sensitivity of waveform mismatch to waveform deformations. We use stationary Gaussian noise with the Advanced LIGO design spectrum for 262 sources and a measured Hanford spectrum from the third observing run (O3) for 132 sources, and also test injections into recorded Hanford strain. On the two Gaussian datasets, the network identifies the exponent of deviations clearly above the noise with accuracies of $0.427$ and $0.329$. To interpret these results, we compare the network with an approximate classifier built from predicted waveform residuals for the six phase families and same sources. The two classifiers often confuse the same families. Examining residuals left by different phase corrections, we link these errors to similarities between the underlying deviations. On selected recorded Hanford strain, networks trained on simulated or recorded noise achieve overall classification accuracies close to those on simulated Hanford noise. Of thirteen neural-network variants tested on the Advanced LIGO design dataset, the recurrent network has the highest mean classification accuracy for loud deviations, although all variants remain below the approximate classifier. The network nearly matches that classifier's accuracy, indicating that identification is limited by similarities between neighboring phase families rather than by the network, for known intrinsic source parameters.
发表机构
- Heidelberg University（海德堡大学）
机构由 AI 辅助整理，请以论文原文为准。
