Data Analysis, Statistics and Probability

New submissions

• New submissions
• Cross-lists
• Replacements

New submissions for Fri, 27 Jan 23

[1]  arXiv:2301.10860 [pdf, other]

Title: Skeleton coupling: a novel interlayer mapping of community evolution in temporal networks

Authors: Bengier Ulgen Kilic, Sarah Feldt Muldoon

Comments: 19 pages, 8 figures

Subjects: Data Analysis, Statistics and Probability (physics.data-an); Adaptation and Self-Organizing Systems (nlin.AO)

Dynamic community detection (DCD) in temporal networks is a complicated task that involves the selection of an algorithm and its associated parameters. How to choose the most appropriate algorithm generally depends on the type of network being analyzed and the specific properties of the data that define the network. In functional temporal networks derived from neuronal spike train data, communities are expected to be transient, and it is common for the network to contain multiple singleton communities. Here, we compare the performance of different DCD algorithms on functional temporal networks built from synthetic neuronal time series data with known community structure. We find that, for these networks, DCD algorithms that utilize interlayer links to perform community carryover between layers outperform other methods. However, we also observe that algorithm performance is highly dependent on the topology of interlayer links, especially in the presence of singleton and transient communities. We therefore define a novel method for defining interlayer links in temporal networks called skeleton coupling that is specifically designed to enhance the linkage of communities in the network throughout time based on the topological properties of the community history. We show that integrating skeleton coupling with current DCD methods improves algorithm performance in synthetic data with planted singleton and transient communities. The use of skeleton coupling to perform DCD will therefore allow for more accurate and interpretable results of community evolution in real-world neuronal data or in other systems with transient structure and singleton communities.

Cross-lists for Fri, 27 Jan 23

[2]  arXiv:2301.11070 (cross-list from physics.chem-ph) [pdf, other]

Title: Additive Eigenvector Analysis of Diffusion: Wave-function Phase as Optimal Reaction Coordinate

Authors: Sergei V. Krivov

Subjects: Chemical Physics (physics.chem-ph); Soft Condensed Matter (cond-mat.soft); Statistical Mechanics (cond-mat.stat-mech); Probability (math.PR); Data Analysis, Statistics and Probability (physics.data-an)

Complex multidimensional stochastic dynamics can be approximately described as a diffusion along reaction coordinates (RCs). If the RCs are optimally selected important properties of dynamics can be computed exactly. The committor is a primary example of an optimal RC. Recently, additive eigenvector (addevs) have been introduced in order to extend the formalism to non-equilibrium dynamics. An addev describes a sub-ensemble of trajectories together with an optimal RC. The sub-ensemble is conditioned to have a single RC optimal for both the forward and time-reversed non-equilibrium dynamics of the sub-ensemble. Here we continue the fundamental development of the framework and illustrate it by analysing diffusion. In particular, we show that the forward and time-reversed committors are functions of the addev, meaning a diffusive model along an addev RC can be used to compute important properties of non-equilibrium dynamics exactly. While the addev conditioning is a reasonable requirement for a RC, it is rather peculiar for physics in general. Nevertheless, we show that one family of addev solutions for diffusion is described by equations of classical mechanics, while another family can be approximated by quantum mechanical equations. We show how the potential can be introduced in to the formalism self-consistently. The developments are illustrated on simple examples. Deep fundamental connections between quantum mechanics and stochastic processes, in particular, between the Schr\"odinger and diffusion equations, are well established. Here we demonstrate yet another one, emphasizing the role of the phase of a wave-function as an optimal RC.

Replacements for Fri, 27 Jan 23

[3]  arXiv:2202.01762 (replaced) [pdf]

Title: Predicting Wind-Driven Spatial Deposition through Simulated Color Images using Deep Autoencoders

Authors: M. Giselle Fernández-Godino, Donald D. Lucas, Qingkai Kong

Comments: 12 pages, 7 figures

Journal-ref: Sci Rep 13, 1394 (2023)

Subjects: Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an); Geophysics (physics.geo-ph)

• New submissions
• Cross-lists
• Replacements