-
Engineered Ordinary Differential Equations as Classification Algorithm (EODECA): thorough characterization and testing
Authors:
Raffaele Marino,
Lorenzo Buffoni,
Lorenzo Chicchi,
Lorenzo Giambagli,
Duccio Fanelli
Abstract:
EODECA (Engineered Ordinary Differential Equations as Classification Algorithm) is a novel approach at the intersection of machine learning and dynamical systems theory, presenting a unique framework for classification tasks [1]. This method stands out with its dynamical system structure, utilizing ordinary differential equations (ODEs) to efficiently handle complex classification challenges. The…
▽ More
EODECA (Engineered Ordinary Differential Equations as Classification Algorithm) is a novel approach at the intersection of machine learning and dynamical systems theory, presenting a unique framework for classification tasks [1]. This method stands out with its dynamical system structure, utilizing ordinary differential equations (ODEs) to efficiently handle complex classification challenges. The paper delves into EODECA's dynamical properties, emphasizing its resilience against random perturbations and robust performance across various classification scenarios. Notably, EODECA's design incorporates the ability to embed stable attractors in the phase space, enhancing reliability and allowing for reversible dynamics. In this paper, we carry out a comprehensive analysis by expanding on the work [1], and employing a Euler discretization scheme. In particular, we evaluate EODECA's performance across five distinct classification problems, examining its adaptability and efficiency. Significantly, we demonstrate EODECA's effectiveness on the MNIST and Fashion MNIST datasets, achieving impressive accuracies of $98.06\%$ and $88.21\%$, respectively. These results are comparable to those of a multi-layer perceptron (MLP), underscoring EODECA's potential in complex data processing tasks. We further explore the model's learning journey, assessing its evolution in both pre and post training environments and highlighting its ability to navigate towards stable attractors. The study also investigates the invertibility of EODECA, shedding light on its decision-making processes and internal workings. This paper presents a significant step towards a more transparent and robust machine learning paradigm, bridging the gap between machine learning algorithms and dynamical systems methodologies.
△ Less
Submitted 20 May, 2024; v1 submitted 22 December, 2023;
originally announced December 2023.
-
Direct observational evidence of an oceanic dual kinetic energy cascade and its seasonality
Authors:
Dhruv Balwada,
Jin-Han Xie,
Raffaele Marino,
Fabio Feraco
Abstract:
The Ocean's turbulent energy cycle has a paradox; large-scale eddies under the control of Earth's rotation primarily transfer kinetic energy (KE) to larger scales via an inverse cascade, while a transfer to smaller scales is needed to accomplish dissipation. It has been argued, using numerical simulations, that fronts, waves and other turbulent structures can produce a forward cascade of KE toward…
▽ More
The Ocean's turbulent energy cycle has a paradox; large-scale eddies under the control of Earth's rotation primarily transfer kinetic energy (KE) to larger scales via an inverse cascade, while a transfer to smaller scales is needed to accomplish dissipation. It has been argued, using numerical simulations, that fronts, waves and other turbulent structures can produce a forward cascade of KE toward dissipation scales. However, this forward cascade and its coexistence with known inverse cascade were not confirmed in observations. Here we present the first evidence of a dual KE cascade in the Ocean by analyzing velocity measurements from surface drifters released in the Gulf of Mexico. Our results show that KE is injected at two dominant scales and transferred to both large and small scales, with the downscale flux dominating at scales smaller than ~1-10km. The cascade rates are modulated seasonally, with stronger KE injection and forward transfer during winter.
△ Less
Submitted 17 February, 2022;
originally announced February 2022.
-
Chaos in Wavy-Stratified Fluid-Fluid Flow
Authors:
Avinash Vaidheeswaran,
Alejandro Clausse,
William D. Fullmer,
Raul Marino,
Martin Lopez de Bertodano
Abstract:
We perform a non-linear analysis of a fluid-fluid wavy-stratified flow using a simplified two-fluid model, i.e., the fixed-flux model (FFM) which is an adaptation of shallow water theory for the two-layer problem. Linear analysis using the perturbation method illustrates the short-wave physics leading to the Kelvin-Helmholtz instability (KHI). The interface dynamics are chaotic and analysis beyond…
▽ More
We perform a non-linear analysis of a fluid-fluid wavy-stratified flow using a simplified two-fluid model, i.e., the fixed-flux model (FFM) which is an adaptation of shallow water theory for the two-layer problem. Linear analysis using the perturbation method illustrates the short-wave physics leading to the Kelvin-Helmholtz instability (KHI). The interface dynamics are chaotic and analysis beyond the onset of instability is required to understand the non-linear evolution of waves. The two-equation FFM solver based on a higher-order spatio-temporal finite difference discretization scheme is used in the current simulations. The solution methodology is verified and the results are compared with the measurements from a laboratory-scale experiment. The Finite-Time Lyapunov Exponent (FTLE) based on simulations is comparable and slightly higher than the Autocorrelation function (ACF) decay rate, consistent with findings from previous studies. Furthermore, the FTLE is observed to be a strong function of the angle of inclination, while the root mean square (RMS) of the interface height exhibits a square-root dependence. It is demonstrated that this simple 1-D FFM captures the essential chaotic features of the interfacial behavior.
△ Less
Submitted 26 September, 2018;
originally announced September 2018.
-
Geophysical turbulence and the duality of the energy flow across scales
Authors:
A. Pouquet,
R. Marino
Abstract:
The ocean and the atmosphere, and hence the climate, are governed at large scale by interactions between pressure gradient, Coriolis and buoyancy forces. This leads to a quasi-geostrophic balance in which, in a two-dimensional-like fashion, the energy injected by solar radiation, winds or tides goes to large scales in what is known as an inverse cascade. Yet, except for Ekman friction, energy diss…
▽ More
The ocean and the atmosphere, and hence the climate, are governed at large scale by interactions between pressure gradient, Coriolis and buoyancy forces. This leads to a quasi-geostrophic balance in which, in a two-dimensional-like fashion, the energy injected by solar radiation, winds or tides goes to large scales in what is known as an inverse cascade. Yet, except for Ekman friction, energy dissipation and turbulent mixing occur at small scale implying the formation of such scales associated with breaking of geostrophic dynamics through wave-eddy interactions \cite{ledwell_00, vanneste_13} or frontogenesis \cite{hoskins_72, mcwilliams_10}, in opposition to the inverse cascade. Can it be both at the same time? We exemplify here this dual behavior of energy with the help of three-dimensional direct numerical simulations of rotating stratified Boussinesq turbulence. We show that efficient small-scale mixing and large-scale coherence develop simultaneously in such geophysical and astrophysical flows, both with constant flux as required by theoretical arguments, thereby clearly resolving the aforementioned contradiction.
△ Less
Submitted 29 September, 2013; v1 submitted 10 September, 2013;
originally announced September 2013.
-
Observation of inertial energy cascade in interplanetary space plasma
Authors:
Luca Sorriso-Valvo,
Raffaele Marino,
Vincenzo Carbone,
Fabio Lepreti,
Pierluigi Veltri,
Alain Noullez,
Roberto Bruno,
Bruno Bavassano,
Ermanno Pietropaolo
Abstract:
We show in this article direct evidence for the presence of an inertial energy cascade, the most characteristic signature of hydromagnetic turbulence (MHD), in the solar wind as observed by the Ulysses spacecraft. After a brief rederivation of the equivalent of Yaglom's law for MHD turbulence, we show that a linear relation is indeed observed for the scaling of mixed third order structure functi…
▽ More
We show in this article direct evidence for the presence of an inertial energy cascade, the most characteristic signature of hydromagnetic turbulence (MHD), in the solar wind as observed by the Ulysses spacecraft. After a brief rederivation of the equivalent of Yaglom's law for MHD turbulence, we show that a linear relation is indeed observed for the scaling of mixed third order structure functions involving Elsässer variables. This experimental result, confirming the prescription stemming from a theorem for MHD turbulence, firmly establishes the turbulent character of low-frequency velocity and magnetic field fluctuations in the solar wind plasma.
△ Less
Submitted 9 February, 2007;
originally announced February 2007.
