Classification of anomalous diffusion in animal movement data using power spectral analysis

Journal of Physics A Mathematical and Theoretical, Journal Year: 2022, Volume and Issue: 55(33), P. 334004 - 334004

Published: July 5, 2022

The field of movement ecology has seen a rapid increase in high-resolution data recent years, leading to the development numerous statistical and numerical methods analyse relocation trajectories. Data are often collected at level individual for long periods that may encompass range behaviours. Here, we use power spectral density (PSD) characterise random patterns black-winged kite (Elanus caeruleus) white stork (Ciconia ciconia). tracks first segmented clustered into different behaviours (movement modes), each mode measure PSD ageing properties process. For foraging find $1/f$ noise, previously reported ecological systems mainly context population dynamics, but not data. We further suggest plausible models behavioural modes by comparing both measured exponents distribution single-trajectory known theoretical results simulations.

Language: Английский

---

Non-Gaussian displacement distributions in models of heterogeneous active particle dynamics

New Journal of Physics, Journal Year: 2023, Volume and Issue: 25(1), P. 013010 - 013010

Published: Jan. 1, 2023

We study the effect of randomly distributed diffusivities and speeds in two models for active particle dynamics with passive fluctuations. demonstrate how non-Gaussian displacement distributions emerge these long time limit, including Cauchy-type exponential (Laplace) shapes. Notably resulting shapes considered here are striking contrast to diffusion models. For motion our discussion points out differences between active- passive-noise. Specifically, we that case active-noise is nice agreement measured data distribution social amoeba.

Language: Английский

---

Ergodic characterization of nonergodic anomalous diffusion processes

Physical Review Research, Journal Year: 2023, Volume and Issue: 5(2)

Published: May 31, 2023

Anomalous diffusion in various complex systems abounds nature and spans multiple space time scales. Canonical characterization techniques that rely upon mean squared displacement break down for nonergodic processes, making it challenging to characterize anomalous from an individual time-series measurement. Nonergodicity reigns when the time-averaged square differs ensemble-averaged even limit of long measurement series. In these cases, typical theoretical results ensemble averages cannot be used understand interpret data acquired averages. The difficulty then lies obtaining statistical descriptors measured process are not nonergodic. We show linear such as standard deviation, coefficient variation, root ergodicity proportion nonergodicity process. contrast, series addressing sequential structure its potential nonlinearity: multifractality change a time-independent way fulfill ergodic assumption, largely independent series' nonergodicity. findings follow multiplicative cascades underlying processes. Adding fractal multifractal would improve Two particular points bear emphasis here. First, appropriate formalism encoding nonlinearity might generate nonergodicity, modeling offers can behave ergodically enough meet needs modeling. Second, this capacity describe processes terms possibility could unify several disparate into common framework.

Language: Английский

---

Minimal model of diffusion with time changing Hurst exponent

Journal of Physics A Mathematical and Theoretical, Journal Year: 2023, Volume and Issue: 56(35), P. 35LT01 - 35LT01

Published: Aug. 2, 2023

Abstract We introduce the stochastic process of incremental multifractional Brownian motion (IMFBM), which locally behaves like fractional with a given local Hurst exponent and diffusivity. When these parameters change as function time responds to evolution gradually: only new increments are governed by parameters, while still retaining power-law dependence on past process. obtain mean squared displacement correlations IMFBM elementary formulas. also provide comparison simulations estimation methods for IMFBM. This mathematically simple is useful in description anomalous diffusion dynamics changing environments, e.g. viscoelastic systems, or when an actively moving particle changes its degree persistence mobility.

Language: Английский

---

Multifractal spectral features enhance classification of anomalous diffusion

Physical review. E, Journal Year: 2024, Volume and Issue: 109(4)

Published: April 15, 2024

Anomalous diffusion processes, characterized by their nonstandard scaling of the mean-squared displacement, pose a unique challenge in classification and characterization. In previous study [Mangalam et al., Phys. Rev. Res. 5, 023144 (2023)], we established comprehensive framework for understanding anomalous using multifractal formalism. The present delves into potential spectral features effectively distinguishing trajectories from five widely used models: fractional Brownian motion, scaled continuous-time random walk, annealed transient time L\'evy walk. We generate extensive datasets comprising ${10}^{6}$ these models extract multiple spectra each trajectory to accomplish this. Our investigation entails thorough analysis neural network performance, encompassing derived varying numbers spectra. also explore integration traditional feature datasets, enabling us assess impact comprehensively. To ensure statistically meaningful comparison, categorize concept groups train networks designated group. Notably, several demonstrate similar levels accuracy, with highest performance observed utilizing moving-window characteristics $p$ varation features. Multifractal features, particularly those three involving different timescales cutoffs, closely follow, highlighting robust discriminatory potential. Remarkably, exclusively trained on single spectrum exhibits commendable surpassing other groups. summary, our findings underscore diverse potent efficacy enhancing predictive capacity machine learning classify processes.

Language: Английский

---

Anomalous diffusion of active Brownian particles in responsive elastic gels: Nonergodicity, non-Gaussianity, and distributions of trapping times

Physical review. E, Journal Year: 2024, Volume and Issue: 110(4)

Published: Oct. 29, 2024

Understanding actual transport mechanisms of self-propelled particles (SPPs) in complex elastic gels---such as the cell cytoplasm, vitro networks chromatin or F-actin fibers, mucus gels---has far-reaching consequences. Implications beyond biology/biophysics are engineering and medicine, with a particular focus on microrheology targeted drug delivery. Here, we examine via extensive computer simulations dynamics SPPs deformable gellike structures responsive to thermal fluctuations. We treat tracer comparable larger than mesh size gel. observe distinct trapping events active tracers at relatively short times, leading subdiffusion; it is followed by an escape from meshwork-induced traps due flexibility network, resulting superdiffusion. thus find crossovers between different regimes. also pronounced nonergodicity non-Gaussianity intermediate times. The distributions times escaping ``cages'' our quasiperiodic gel often reveal existence two timescales dynamics. At high activity these become comparable. Furthermore, that mean waiting time exhibits power-law dependence (in terms their P\'eclet number). Our results additionally showcase both exponential nonexponential activities. Extensions this setup possible, factors such anisotropy particles, topologies various interactions (also nonlocal nature) be considered.

Language: Английский

---

Extracting, quantifying, and comparing dynamical and biomechanical properties of living matter through single particle tracking

Physical Chemistry Chemical Physics, Journal Year: 2022, Volume and Issue: 25(3), P. 1513 - 1537

Published: Dec. 22, 2022

A panoply of new tools for tracking single particles and molecules has led to novel insights into physical properties living matter governing cellular development function, health disease.

Language: Английский

---

Modelling anomalous diffusion in semi-infinite disordered systems and porous media

New Journal of Physics, Journal Year: 2022, Volume and Issue: 24(12), P. 123004 - 123004

Published: Nov. 30, 2022

Abstract For an effectively one-dimensional, semi-infinite disordered system connected to a reservoir of tracer particles kept at constant concentration, we provide the dynamics concentration profile. Technically, start with Montroll–Weiss equation continuous time random walk scale-free waiting density. From this pass formulation in terms fractional diffusion for profile C(x,t) space boundary condition $C(0,t) = C_0$?> stretchy="false">(0,t)=C0 , using subordination approach. deduce flux and so-called breakthrough curve (BTC) given distance from source. In particular, BTCs are routinely measured geophysical contexts but also interest single-particle tracking experiments. ‘residual’ BTCs, by $1-P(x,t)$?> overflow="scroll">1−P

Language: Английский

---

Gramian angular fields for leveraging pretrained computer vision models with anomalous diffusion trajectories

Physical review. E, Journal Year: 2023, Volume and Issue: 107(3)

Published: March 28, 2023

Anomalous diffusion is present at all scales, from atomic to large ones. Some exemplary systems are ultracold atoms, telomeres in the nucleus of cells, moisture transport cement-based materials, arthropods' free movement, and birds' migration patterns. The characterization gives critical information about dynamics these provides an interdisciplinary framework with which study diffusive transport. Thus, problem identifying underlying regimes inferring anomalous exponent α high confidence physics, chemistry, biology, ecology. Classification analysis raw trajectories combining machine learning techniques statistics extracted them have widely been studied Diffusion Challenge [Muñoz-Gil et al., Nat. Commun. 12, 6253 (2021)2041-172310.1038/s41467-021-26320-w]. Here we a new data-driven method for working trajectories. This utilizes Gramian angular fields (GAF) encode one-dimensional as images (Gramian matrices), while preserving their spatiotemporal structure input computer-vision models. allows us leverage two well-established pretrained models, ResNet MobileNet, characterize regime infer α. Short lengths between 10 50 commonly encountered single-particle tracking experiments most difficult ones characterize. We show that GAF can outperform current state-of-the-art increasing accessibility methods applied setting.

Language: Английский

---

Modelling intermittent anomalous diffusion with switching fractional Brownian motion

New Journal of Physics, Journal Year: 2023, Volume and Issue: 25(10), P. 103031 - 103031

Published: Oct. 1, 2023

Abstract The stochastic trajectories of molecules in living cells, as well the dynamics many other complex systems, often exhibit memory their path over long periods time. In addition, these systems can show dynamic heterogeneities due to which motion changes along trajectories. Such effects manifest themselves spatiotemporal correlations. Despite broad occurrence heterogeneous nature, analysis is still quite poorly understood and tools model them are largely missing. We contribute tackling this problem by employing an integral representation Mandelbrot’s fractional Brownian that compliant with varying parameters while maintaining memory. Two types switching analysed, transitions arising from a Markovian process scale-free intermittent processes. obtain simple formulas for classical statistics processes, namely mean squared displacement power spectral density. Further, method identify based on distribution displacements described. A validation given experimental measurements quantum dots cytoplasm live mammalian cells were obtained single-particle tracking.

Language: Английский

---