Application of symbolic recurrence to experimental data, from firearm prevalence to fish swimming
Application of symbolic recurrence to experimental data, from firearm prevalence to fish swimming
复制标题
DOI：
10.1063/1.5119883
复制
发表时间：
2019-11-01
期刊：
影响因子：
2.9
通讯作者：
Porfiri, Maurizio
中科院分区：
文献类型：
--
作者：
Boldini, Alain;Karakaya, Mert;Porfiri, Maurizio
递归图和递归量化分析是研究动力系统行为的有力工具。我们通过这些工具学习的内容通常由相空间中距离阈值的选择决定，这在递归的定义中引入了任意性。符号递归不仅克服了这个困难，而且它提供了一个更丰富的表示，保留了相空间的递归部分。使用符号递归，我们可以构建递归图，进行量化分析，并从它们的时间序列中检查动力系统之间的因果关系。虽然以前的努力已经证明了这种符号框架的合成数据的可行性，真实的时间序列的研究仍然难以捉摸。在这里，我们试图通过系统地研究广泛的实验数据集来弥合这一差距，从美国的枪支流行率和媒体报道到性别对游泳鱼相互作用的影响。这项工作提供了一个令人信服的演示符号递归在不同研究领域的现实世界的应用研究的潜力，同时提供了一个计算机代码，为研究人员进行自己的时间序列探索。由AIP Publishing授权出版。
Recurrence plots and recurrence quantification analysis are powerful tools to study the behavior of dynamical systems. What we learn through these tools is typically determined by the choice of a distance threshold in the phase space, which introduces arbitrariness in the definition of recurrence. Not only does symbolic recurrence overcome this difficulty, but also it offers a richer representation that book-keeps the recurrent portions of the phase space. Using symbolic recurrences, we can construct recurrence plots, perform quantification analysis, and examine causal links between dynamical systems from their time-series. Although previous efforts have demonstrated the feasibility of such a symbolic framework on synthetic data, the study of real time-series remains elusive. Here, we seek to bridge this gap by systematically examining a wide range of experimental datasets, from firearm prevalence and media coverage in the United States to the effect of sex on the interaction of swimming fish. This work offers a compelling demonstration of the potential of symbolic recurrence in the study of real-world applications across different research fields while providing a computer code for researchers to perform their own time-series explorations. Published under license by AIP Publishing.