arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2110.03533physics.data-annlin.CD
收藏

NoLiTiA:一个用于非线性时间序列分析的开源工具

NoLiTiA: An Open-Source Toolbox for Nonlinear Time Series Analysis

Immo Weber, Carina Renate Oehrn

摘要:

在许多科学领域,如神经科学,气候学或物理学,复杂的关系可以用非线性力学来描述。尽管它们的相关性,许多科学家仍然应用线性估计,以评估复杂的相互作用。这部分是由于缺乏非线性方法的综合汇编。可用的软件包大多只专注于非线性时间序列分析的一个方面,并且通常需要一些编码能力才能使用。在这里,我们介绍NoLiTiA,一个用于非线性时间序列分析的免费开源MATLAB工具箱。与其他目前可用的非线性软件包相比,NoLiTiA提供了1)广泛的经典和最近开发的方法的实现,2)新提出的空间和时间分辨递归分析的实现,以及3)由于图形用户界面和批处理编辑器,即使对几乎没有编码经验的用户也可以访问的直观环境。其核心方法来自复杂系统理论的三个不同领域,包括动力系统理论、递归量化分析和信息论。除了既定的方法,包括估计动态不变量,如李雅普诺夫指数和基于熵的措施,如主动信息存储,我们包括最近的事态发展量化的时间分辨非周期振荡。总的来说,该工具箱将使从事时间序列处理的广大科学界能够使用非线性方法。

英文摘要:

In many scientific fields like e.g. neuroscience, climatology or physics, complex relationships can be described most parsimoniously by nonlinear mechanics. Despite their relevance, many scientists still apply linear estimates in order to evaluate complex interactions. This is partially due to the lack of a comprehensive compilation of nonlinear methods. Available packages mostly specialize in only one aspect of nonlinear time-series analysis and most often require some coding proficiency to use. Here, we introduce NoLiTiA, a free open-source MATLAB toolbox for nonlinear time series analysis. In comparison to other currently available nonlinear packages, NoLiTiA offers 1) an implementation of a broad range of classic and recently developed methods, 2) an implementation of newly proposed spatially and time-resolved recurrence analysis and 3) an intuitive environment accessible even to users with little coding experience due to a graphical user interface and batch-editor. The core methodology derives from three distinct fields of complex systems theory, including dynamical systems theory, recurrence quantification analysis and information theory. Besides established methodology including estimation of dynamic invariants like Lyapunov exponents and entropy-based measures, such as active information storage, we include recent developments of quantifying time-resolved aperiodic oscillations. In general, the toolbox will make nonlinear methods accessible to the broad scientific community engaged in time series processing.

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。

↑