arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1709.05042nlin.CD
收藏

用于时间序列数据分析的度加权递归网络

Degree weighted recurrence networks for the analysis of time series data

Rinku Jacob, K. P. Harikrishnan, R. Misra, G. Ambika

摘要:

递归网络是一种有效的时间序列数据非线性分析的有力工具。在这种情况下的分析主要是做与非加权和无向复杂网络构造的特定标准的时间序列。在这项工作中,我们提出了一种从时间序列构建“加权递归网络”(WRN)的新方法,并展示了它如何揭示有关混沌吸引子结构的有用信息,而通常的未加权递归网络无法提供这些信息。特别地,我们发现WRN的节点强度分布,从每一个混沌吸引子遵循幂律(指数尾)的指数特性的吸引子的分形结构。这导致了复杂网络中的一个新类别,所有标准混沌吸引子的网络都属于这个类别。此外,我们提出了聚类系数和特征路径长度的广义定义,并表明这些措施可以有效地区分混沌动力学的白噪声和$1/f$有色噪声。我们的研究结果表明,WRN和相关的措施可以成为潜在的重要工具,从现实世界的系统,因为它们是明确区分的噪声或随机系统的短和嘈杂的时间序列的分析。

英文摘要:

Recurrence networks are powerful tools used effectively in the nonlinear analysis of time series data. The analysis in this context is done mostly with unweighted and undirected complex networks constructed with specific criteria from the time series. In this work, we propose a novel method to construct "weighted recurrence network"(WRN) from a time series and show how it can reveal useful information regarding the structure of a chaotic attractor, which the usual unweighted recurrence network cannot provide. Especially, we find the node strength distribution of the WRN, from every chaotic attractor follows a power law (with exponential tail) with the index characteristic to the fractal structure of the attractor. This leads to a new class among complex networks, to which networks from all standard chaotic attractors are found to belong. In addition, we present generalized definitions for clustering coefficient and characteristic path length and show that these measures can effectively discriminate chaotic dynamics from white noise and $1/f$ colored noise. Our results indicate that the WRN and the associated measures can become potentially important tools for the analysis of short and noisy time series from the real world systems as they are clearly demarked from that of noisy or stochastic systems.

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。

↑