Parallelization of a Common Changepoint Detection Method

Parallelization of a Common Changepoint Detection Method
复制标题

常见变点检测方法的并行化

DOI:
10.1080/10618600.2019.1647216
复制
发表时间:
2019
影响因子:
2.4
通讯作者:
Tickle S
Tickle S
中科院分区:
数学2区
文献类型:
--
作者:
Tickle S

文献摘要

参考文献

被引文献

相似文献

近年来,已经提出了各种有效检测变点的方法,其中一种流行的方法涉及使用动态规划来最小化惩罚成本函数。在某些情况下,这些算法可以具有在数据点的数量上是线性的预期计算成本;然而,最坏情况下的成本仍然是二次的。我们介绍了两种提高这些方法的计算性能的方法,都是基于并行化的动态规划方法。我们建立了并行化可以提供大量的计算改进:在某些情况下,计算成本大致以所使用的核心数量的二次方下降。这些并行实现不再保证找到真正的最小惩罚成本,但是,我们表明,他们保留了相同的渐近保证,在估计的数量和位置的变化的准确性。本文的补充材料可在网上查阅。
In recent years, various means of efficiently detecting changepoints have been proposed, with one popular approach involving minimizing a penalized cost function using dynamic programming. In some situations, these algorithms can have an expected computational cost that is linear in the number of data points; however, the worst case cost remains quadratic. We introduce two means of improving the computational performance of these methods, both based on parallelizing the dynamic programming approach. We establish that parallelization can give substantial computational improvements: in some situations the computational cost decreases roughly quadratically in the number of cores used. These parallel implementations are no longer guaranteed to find the true minimum of the penalized cost; however, we show that they retain the same asymptotic guarantees in terms of their accuracy in estimating the number and location of the changes. Supplementary materials for this article are available online.
变化点检测的惩罚学习
DOI: --
发表时间: 2017
期刊: European Signal Processing Conference
影响因子: --
作者:
Charles Truong;L. Oudre;N. Vayatis
通讯作者: N. Vayatis
DOI: 10.1093/biostatistics/kxh008
发表时间: 2004-10-01
期刊: BIOSTATISTICS
影响因子: 2.1
作者:
Olshen, AB;Venkatraman, ES;Wigler, M
通讯作者: Wigler, M
DOI: --
发表时间: 2017
期刊:
影响因子: --
作者:
Fuqi Chen;S. Nkurunziza
通讯作者: S. Nkurunziza
DOI: 10.1093/bioinformatics/bti500
发表时间: 2005-08-01
期刊: BIOINFORMATICS
影响因子: 5.8
作者:
Margolin, AA;Greshock, J;Weber, BL
通讯作者: Weber, BL