Gravitational-wave signal recognition of LIGO data by deep learning
Gravitational-wave signal recognition of LIGO data by deep learning
复制标题
DOI：
10.1103/physrevd.101.104003
复制
发表时间：
2019-09
影响因子：
5
通讯作者：
He Wang;Shichao Wu;Zhoujian Cao;Xiaolin Liu;Jianhua Zhu
中科院分区：
文献类型：
--
作者：
He Wang;Shichao Wu;Zhoujian Cao;Xiaolin Liu;Jianhua Zhu
深度学习方法作为数据分析的工具近年来发展非常快。这种技术在处理引力波探测数据方面非常有前景。文献中已经有许多使用深度学习技术来处理模拟引力波数据的工作。在本文中，我们将深度学习应用于 LIGO O1 数据。为了提高弱信号识别能力，我们对卷积神经网络（CNN）进行了一些调整。我们调整后的卷积神经网络的信号识别精度和效率与文献中发表的其他深度学习作品相当。基于我们调整后的CNN，我们可以清楚地识别O1和O2中包含的11个已确认的引力波事件。我们在 O1 数据中发现了大约 2000 个引力波触发点。
Deep learning method develops very fast as a tool for data analysis these years. Such a technique is quite promising to treat gravitational wave detection data. There are many works already in the literature which used deep learning technique to process simulated gravitational wave data. In this paper we apply deep learning to LIGO O1 data. In order to improve the weak signal recognition we adjust the convolutional neural network (CNN) a little bit. Our adjusted convolutional neural network admits comparable accuracy and efficiency of signal recognition as other deep learning works published in the literature. Based on our adjusted CNN, we can clearly recognize the eleven confirmed gravitational wave events included in O1 and O2. And more we find about 2000 gravitational wave triggers in O1 data.