A nonparametric method to evaluate significance of events in search for gravitational waves with false discovery rate
A nonparametric method to evaluate significance of events in search for gravitational waves with false discovery rate
复制标题
DOI：
10.1088/1742-6596/1468/1/012222
复制
发表时间：
2019-06
期刊：
影响因子：
--
通讯作者：
H. Yuzurihara;S. Mano;H. Tagoshi
中科院分区：
文献类型：
--
作者：
H. Yuzurihara;S. Mano;H. Tagoshi
在LIGO和Virgo最近的观测运行中，已经观测到数十个引力波事件。LIGO/Virgo/KAGRA的引力波网络计划在未来进行更长时间的观测，并提高灵敏度。在这种情况下，使用可靠的方法评估每个事件的重要性非常重要，因为我们预计，随着观察期变长以及检测器灵敏度的提高，我们预计会在给定阈值的情况下检测到更多的假阳性事件。误报率。本文引入了q值，q值是Storey在文献[1]中首次提出的，并被解释为错误发现率（FDR）。我们介绍了一种算法来评估FDR在一个非参数的方式，并将其应用到先进的LIGO O 1数据的分析，并发现我们可以获得一个更好的显着性GW 151012具有最低的O 1事件。对于其他事件，我们获得了LIGO和Virgo已经获得的一致结果。
During recent observation runs by LIGO and Virgo, tens of gravitational wave events have been observed. The gravitational wave network of LIGO/Virgo/KAGRA is planning longer observation runs with improved sensitivity in the future. In this situation, it is important to evaluate the significance of each event with reliable methods since we expect, as the observation period becomes longer and as the sensitivity of the detector improves, we expect to detect more false positive events with a given threshold for the false alarm rate. In this paper, we introduce q-value which was first introduced by Storey[1], and is interpreted as False Discovery Rate (FDR). We introduce an algorithm to evaluate FDR in a non-parametric way, and apply it to publically available results of the analysis of the advanced LIGO O1 data, and find that we can obtain a better significance for GW151012 which has the lowest significance in the O1 events. For other events, we obtain consistent results already obtained by LIGO and Virgo.