Showing 1–50 of 110 results for author: Tran, M

Control and optimization for Neural Partial Differential Equations in Supervised Learning

Authors: Alain Bensoussan, Minh-Binh Tran, Bangjie Wang

Abstract: Although there is a substantial body of literature on control and optimization problems for parabolic and hyperbolic systems, the specific problem of controlling and optimizing the coefficients of the associated operators within such systems has not yet been thoroughly explored. In this work, we aim to initiate a line of research in control theory focused on optimizing and controlling the coeffici… ▽ More Although there is a substantial body of literature on control and optimization problems for parabolic and hyperbolic systems, the specific problem of controlling and optimizing the coefficients of the associated operators within such systems has not yet been thoroughly explored. In this work, we aim to initiate a line of research in control theory focused on optimizing and controlling the coefficients of these operators-a problem that naturally arises in the context of neural networks and supervised learning. In supervised learning, the primary objective is to transport initial data toward target data through the layers of a neural network. We propose a novel perspective: neural networks can be interpreted as partial differential equations (PDEs). From this viewpoint, the control problem traditionally studied in the context of ordinary differential equations (ODEs) is reformulated as a control problem for PDEs, specifically targeting the optimization and control of coefficients in parabolic and hyperbolic operators. To the best of our knowledge, this specific problem has not yet been systematically addressed in the control theory of PDEs. To this end, we propose a dual system formulation for the control and optimization problem associated with parabolic PDEs, laying the groundwork for the development of efficient numerical schemes in future research. We also provide a theoretical proof showing that the control and optimization problem for parabolic PDEs admits minimizers. Finally, we investigate the control problem associated with hyperbolic PDEs and prove the existence of solutions for a corresponding approximated control problem. △ Less

Submitted 25 June, 2025; originally announced June 2025.

A large scaling property of level sets for degenerate $p$-Laplacian equations with logarithmic BMO matrix weights

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: In this study, we deal with generalized regularity properties for solutions to $p$-Laplace equations with degenerate matrix weights. It has already been observed in previous interesting works [A. Kh. Balci, L. Diening, R. Giova, A. Passarelli di Napoli, SIAM J. Math. Anal. 54(2022), 2373-2412] and [A. Kh. Balci, S.-S. Byun, L. Diening, H.-S. Lee, J. Math. Pures Appl. (9) 177(2023), 484-530] that g… ▽ More In this study, we deal with generalized regularity properties for solutions to $p$-Laplace equations with degenerate matrix weights. It has already been observed in previous interesting works [A. Kh. Balci, L. Diening, R. Giova, A. Passarelli di Napoli, SIAM J. Math. Anal. 54(2022), 2373-2412] and [A. Kh. Balci, S.-S. Byun, L. Diening, H.-S. Lee, J. Math. Pures Appl. (9) 177(2023), 484-530] that gaining Calderón-Zygmund estimates for nonlinear equations with degenerate weights under the so-called $\log$-$\mathrm{BMO}$ condition and minimal regularity assumption on the boundary. In this paper, we also follow this direction and extend general gradient estimates for level sets of the gradient of solutions up to more subtle function spaces. In particular, we construct a covering of the super-level sets of the spatial gradient $|\nabla u|$ with respect to a large scaling parameter via fractional maximal operators. △ Less

Submitted 31 May, 2025; originally announced June 2025.

Comments: 23 pages

Signatures of Type $A$ Root Systems

Authors: Michael Cuntz, Hung Manh Tran, Tan Nhat Tran, Shuhei Tsujie

Abstract: Given a type $A$ root system $Φ$ of rank $n$, we introduce the concept of a signature for each subset $S$ of $Φ$ consisting of $n+1$ positive roots. For a subset $S$ represented by a tuple $(β_1, \ldots, β_{n+1})$, the signature of $S$ is defined as an unordered pair $\{a, b\}$, where $a$ and $b$ denote the numbers of $1$s and $-1$s, respectively, among the cofactors… ▽ More Given a type $A$ root system $Φ$ of rank $n$, we introduce the concept of a signature for each subset $S$ of $Φ$ consisting of $n+1$ positive roots. For a subset $S$ represented by a tuple $(β_1, \ldots, β_{n+1})$, the signature of $S$ is defined as an unordered pair $\{a, b\}$, where $a$ and $b$ denote the numbers of $1$s and $-1$s, respectively, among the cofactors $(-1)^k \det(S \setminus \{β_k\})$ for $1 \le k \le n+1$. We prove that the number of tuples with a given signature can be expressed in terms of classical Eulerian numbers. The study of these signatures is motivated by their connections to the arithmetic and combinatorial properties of cones over deformed arrangements defined by $Φ$, including the Shi, Catalan, Linial, and Ish arrangements. We apply our main result to compute two important invariants of these arrangements: The minimum period of the characteristic quasi-polynomial, and the evaluation of the classical and arithmetic Tutte polynomials at $(1, 1)$. △ Less

Submitted 7 April, 2025; originally announced April 2025.

Comments: 17 pages, 2 figures

MSC Class: Primary 05C50; Secondary 52C35

Formation of condensations for non-radial solutions to 3-wave kinetic equations

Authors: Gigliola Staffilani, Minh-Binh Tran

Abstract: We consider in this work a $2$-dimensional $3$-wave kinetic equation describing the dynamics of the thermal cloud outside a Bose-Einstein Condensate. We construct global non-radial mild solutions for the equation. Those mild solutions are the summation of Dirac masses on circles. We prove that in each spatial direction, either Dirac masses at the origin, which are the so-called Bose-Einstein conde… ▽ More We consider in this work a $2$-dimensional $3$-wave kinetic equation describing the dynamics of the thermal cloud outside a Bose-Einstein Condensate. We construct global non-radial mild solutions for the equation. Those mild solutions are the summation of Dirac masses on circles. We prove that in each spatial direction, either Dirac masses at the origin, which are the so-called Bose-Einstein condensates, can be formed in finite time or the solutions converge to Bose-Einstein condensates as time evolves to infinity. We also describe a dynamics of the formation of the Bose-Einstein condensates latter case. In this case, on each direction, the solutions accumulate around circles close to the origin at growth rates at least linearly in time. △ Less

Submitted 21 March, 2025; originally announced March 2025.

Control, Optimal Transport and Neural Differential Equations in Supervised Learning

Authors: Minh-Nhat Phung, Minh-Binh Tran

Abstract: We study the fundamental computational problem of approximating optimal transport (OT) equations using neural differential equations (Neural ODEs). More specifically, we develop a novel framework for approximating unbalanced optimal transport (UOT) in the continuum using Neural ODEs. By generalizing a discrete UOT problem with Pearson divergence, we constructively design vector fields for Neural O… ▽ More We study the fundamental computational problem of approximating optimal transport (OT) equations using neural differential equations (Neural ODEs). More specifically, we develop a novel framework for approximating unbalanced optimal transport (UOT) in the continuum using Neural ODEs. By generalizing a discrete UOT problem with Pearson divergence, we constructively design vector fields for Neural ODEs that converge to the true UOT dynamics, thereby advancing the mathematical foundations of computational transport and machine learning. To this end, we design a numerical scheme inspired by the Sinkhorn algorithm to solve the corresponding minimization problem and rigorously prove its convergence, providing explicit error estimates. From the obtained numerical solutions, we derive vector fields defining the transport dynamics and construct the corresponding transport equation. Finally, from the numerically obtained transport equation, we construct a neural differential equation whose flow converges to the true transport dynamics in an appropriate limiting regime. △ Less

Submitted 19 May, 2025; v1 submitted 19 March, 2025; originally announced March 2025.

Controlling discrete semilinear wave equations toward flocks

Authors: Sarah Strikwerda, Hung Vinh Tran, Minh-Binh Tran

Abstract: In this work, we initiate the research on controlling nonlinear waves propagating on lattices from a completely new perspective. We consider nonlinear waves on a lattice as a system of interacting particles and study their collective flocking behavior. By designing suitable feedback controls, we show that any admissible flock can be reached within a finite amount of time. Finally, we highlight the… ▽ More In this work, we initiate the research on controlling nonlinear waves propagating on lattices from a completely new perspective. We consider nonlinear waves on a lattice as a system of interacting particles and study their collective flocking behavior. By designing suitable feedback controls, we show that any admissible flock can be reached within a finite amount of time. Finally, we highlight the connection between our flocking problem and a minimal-time problem in the framework of nonlinear Hamilton-Jacobi equations and optimal control theory. △ Less

Submitted 4 June, 2025; v1 submitted 12 February, 2025; originally announced February 2025.

Numerical schemes for a fully nonlinear coagulation-fragmentation model coming from wave kinetic theory

Authors: Arijit Das, Minh-Binh Tran

Abstract: This article introduces a novel numerical approach, based on Finite Volume Techniques, for studying fully nonlinear coagulation-fragmentation models, where both the coagulation and fragmentation components of the collision operator are nonlinear. The models come from $3-$wave kinetic equations, a pivotal framework in wave turbulence theory. Despite the importance of wave turbulence theory in physi… ▽ More This article introduces a novel numerical approach, based on Finite Volume Techniques, for studying fully nonlinear coagulation-fragmentation models, where both the coagulation and fragmentation components of the collision operator are nonlinear. The models come from $3-$wave kinetic equations, a pivotal framework in wave turbulence theory. Despite the importance of wave turbulence theory in physics and mechanics, there have been very few numerical schemes for $3-$wave kinetic equations, in which no ad-hoc additional assumptions are imposed on the evolution of the solutions, and the current manuscript provides one of the first of such schemes. To the best of our knowledge, this also is the first numerical scheme capable of accurately capturing the long-term asymptotic behavior of solutions to a fully nonlinear coagulation-fragmentation model that includes both forward and backward energy cascades. The scheme is implemented on some test problems, demonstrating strong alignment with theoretical predictions of energy cascade rates. We further introduce a weighted Finite Volume variant to ensure energy conservation across varying degrees of kernel homogeneity. Convergence and first-order consistency are established through theoretical analysis and verified by experimental convergence orders in test cases. △ Less

Submitted 3 June, 2025; v1 submitted 6 December, 2024; originally announced December 2024.

MSC Class: 65M08; 45K05; 76F55

Internal Control of The Transition Kernel for Stochastic Lattice Dynamics

Authors: Amirali Hannani, Minh-Nhat Phung, Minh-Binh Tran, Emmanuel Trélat

Abstract: In [5], we have designed impulsive and feedback controls for harmonic chains with a point thermostat. In this work, we study the internal control for stochastic lattice dynamics, with the goal of controlling the transition kernel of the kinetic equation in the limit. A major novelty of the work is the introduction of a new geometric combinatorial argument, used to establish paths for the controls. In [5], we have designed impulsive and feedback controls for harmonic chains with a point thermostat. In this work, we study the internal control for stochastic lattice dynamics, with the goal of controlling the transition kernel of the kinetic equation in the limit. A major novelty of the work is the introduction of a new geometric combinatorial argument, used to establish paths for the controls. △ Less

Submitted 1 October, 2024; v1 submitted 4 July, 2024; originally announced July 2024.

Comments: 39 pages, 1 figure

Numerical Schemes for 3-Wave Kinetic Equations: A Complete Treatment of the Collision Operator

Authors: Steven Walton, Minh-Binh Tran

Abstract: In our previous work, numerical schemes for a simplified version of 3-wave kinetic equations, in which only the simple forward-cascade terms of the collision operators are kept, have been successfully designed, especially to capture the long time dynamics of the equation given the multiple blow-up time phenomenon. In this second work in the series, we propose numerical treatments for the complete… ▽ More In our previous work, numerical schemes for a simplified version of 3-wave kinetic equations, in which only the simple forward-cascade terms of the collision operators are kept, have been successfully designed, especially to capture the long time dynamics of the equation given the multiple blow-up time phenomenon. In this second work in the series, we propose numerical treatments for the complete 3-wave kinetic equations, in which the complete, much more complicated collision operators are fully considered based on a novel conservative form of the equation. We then derive an implicit finite volume scheme to solve the equation. The new discretization uses an adaptive time-stepping method which allows for the simulations to be carried to very long times. Our computed solutions are compared with previously derived long-time asymptotic estimates for the decay rate of total energy of time-dependent solutions of 3-wave kinetic equations and found to be in excellent agreement. △ Less

Submitted 6 June, 2025; v1 submitted 27 February, 2024; originally announced February 2024.

Controlling the Rates of a Chain of Harmonic Oscillators with a Point Langevin Thermostat

Authors: Amirali Hannani, Minh-Binh Tran, Minh Nhat Phung, Emmanuel Trélat

Abstract: We consider the control problem of controlling the rates of an infinite chain of coupled harmonic oscillators with a Langevin thermostat at the origin. We study the effect of two types of open-loop boundary controls, impulsive control and linear memory-feedback control, in the high frequency limit. We investigate their action on the reflection-transmission coefficients for the wave energy for the… ▽ More We consider the control problem of controlling the rates of an infinite chain of coupled harmonic oscillators with a Langevin thermostat at the origin. We study the effect of two types of open-loop boundary controls, impulsive control and linear memory-feedback control, in the high frequency limit. We investigate their action on the reflection-transmission coefficients for the wave energy for the scattering of the thermostat. Our study shows that the impulsive boundary controls have no impact on the rates and are thus not appropriate to act on the system, despite their physical meaning and relevance. In contrast, the second kind of control that we propose, which is less standard and uses the past of the state solution of the system, is adequate and relevant. We prove that any triple of rates satisfying appropriate assumptions is asymptotically reachable thanks to the linear memory-feedback controls that we design explicitly. △ Less

Submitted 17 January, 2025; v1 submitted 12 January, 2024; originally announced January 2024.

A new approach to convergence analysis of iterative models with optimal error bounds

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen, Thai-Hung Nguyen, Tan-Phuc Nguyen, Tien-Khai Nguyen, Cong-Duy-Nguyen Nguyen, Trung-Hieu Huynh

Abstract: In this paper, we study a new approach related to the convergence analysis of Ishikawa-type iterative models to a common fixed point of two non-expansive mappings in Banach spaces. The main novelty of our contribution lies in the so-called \emph{optimal error bounds}, which established some necessary and sufficient conditions for convergence and derived both the error estimates and bounds on the c… ▽ More In this paper, we study a new approach related to the convergence analysis of Ishikawa-type iterative models to a common fixed point of two non-expansive mappings in Banach spaces. The main novelty of our contribution lies in the so-called \emph{optimal error bounds}, which established some necessary and sufficient conditions for convergence and derived both the error estimates and bounds on the convergence rates for iterative schemes. Although a special interest here is devoted to the Ishikawa and modified Ishikawa iterative sequences, the theory of \emph{optimal error bounds} proposed in this paper can also be favorably applied to various types of iterative models to approximate common fixed points of non-expansive mappings. △ Less

Submitted 4 January, 2024; originally announced January 2024.

Comments: 29 pages, 9 figures

Vines and MAT-labeled graphs

Authors: Hung Manh Tran, Tan Nhat Tran, Shuhei Tsujie

Abstract: The present paper explores a connection between two concepts arising from different fields of mathematics. The first concept, called vine, is a graphical model for dependent random variables. This concept first appeared in a work of Joe (1994), and the formal definition was given later by Cooke (1997). Vines have nowadays become an active research area whose applications can be found in probabilit… ▽ More The present paper explores a connection between two concepts arising from different fields of mathematics. The first concept, called vine, is a graphical model for dependent random variables. This concept first appeared in a work of Joe (1994), and the formal definition was given later by Cooke (1997). Vines have nowadays become an active research area whose applications can be found in probability theory and uncertainty analysis. The second concept, called MAT-freeness, is a combinatorial property in the theory of freeness of logarithmic derivation modules of hyperplane arrangements. This concept was first studied by Abe-Barakat-Cuntz-Hoge-Terao (2016), and soon afterwards investigated further by Cuntz-M{ü}cksch (2020). In the particular case of graphic arrangements, the last two authors (2023) recently proved that the MAT-freeness is completely characterized by the existence of certain edge-labeled graphs, called MAT-labeled graphs. In this paper, we first introduce a poset characterization of a vine, the so-called vine. Then we show that, interestingly, there exists an explicit equivalence between the categories of locally regular vines and MAT-labeled graphs. In particular, we obtain an equivalence between the categories of regular vines and MAT-labeled complete graphs. Several applications will be mentioned to illustrate the interaction between the two concepts. Notably, we give an affirmative answer to a question of Cuntz-M{ü}cksch that MAT-freeness can be characterized by a generalization of the root poset in the case of graphic arrangements. △ Less

Submitted 23 February, 2024; v1 submitted 29 November, 2023; originally announced November 2023.

Comments: 32 pages; refined the definitions of the categories MG and LRV (Def. 6.2 & 6.3), hence improved the main result (Thm. 6.10); the term "vineposet" is no longer used, instead we distinguish the graphical and poset definitions of a vine

MSC Class: Primary 06A07; Secondary 52C35

Journal ref: Forum of Mathematics, Sigma 12 (2024) e128

Existence of weak solutions to borderline double-phase problems with logarithmic convection term

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen

Abstract: In this study, we devote our attention to the question of clarifying the existence of a weak solution to a class of quasilinear double-phase elliptic equations with logarithmic convection terms under some appropriate assumptions on data. The proof is based on the surjectivity theorem for the pseudo-monotone operators and modular function spaces and embedding theorems in generalized Orlicz spaces.… ▽ More In this study, we devote our attention to the question of clarifying the existence of a weak solution to a class of quasilinear double-phase elliptic equations with logarithmic convection terms under some appropriate assumptions on data. The proof is based on the surjectivity theorem for the pseudo-monotone operators and modular function spaces and embedding theorems in generalized Orlicz spaces. Our approach in this paper can be extended naturally to a larger class of unbalanced double-phase problems with logarithmic perturbation and gradient dependence on the right-hand sides. △ Less

Submitted 4 January, 2024; v1 submitted 12 September, 2023; originally announced September 2023.

Comments: 23 pages

Regularity for the steady Stokes-type flow of incompressible Newtonian fluids in some generalized function settings

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen, Hong-Nhung Nguyen

Abstract: A study of regularity estimate for weak solution to generalized stationary Stokes-type systems involving $p$-Laplacian is offered. The governing systems of equations are based on steady incompressible flow of a Newtonian fluids. This paper also provides a relatively complete picture of our main results in two regards: problems with nonlinearity is regular with respect to the gradient variable; and… ▽ More A study of regularity estimate for weak solution to generalized stationary Stokes-type systems involving $p$-Laplacian is offered. The governing systems of equations are based on steady incompressible flow of a Newtonian fluids. This paper also provides a relatively complete picture of our main results in two regards: problems with nonlinearity is regular with respect to the gradient variable; and asymtotically regular problems, whose nonlinearity satisfies a particular structure near infinity. For such Stokes-type systems, we derive regularity estimates for both velocity gradient and its associated pressure in two special classes of function spaces: the generalized Lorentz and $ψ$-generalized Morrey spaces. △ Less

Submitted 2 December, 2023; v1 submitted 19 August, 2023; originally announced August 2023.

Comments: 35 pages

Journal ref: Nonlinear Analysis: Real World Applications 77 (2024), 104049

On the wave turbulence theory for the beam wave equation

Authors: Gigliola Staffilani, Minh-Binh Tran

Abstract: Starting from the beam wave equation, which has a Schrödinger structure, on a hypercubic lattice of size $L$, with weak nonlinearity of strength $λ$, we show that the two point correlation function can be asymptotically expressed as the solution of the 3-wave kinetic equation at the kinetic limit under very general assumptions. The limit is taken in the physical order: first taking $L\to\infty$, t… ▽ More Starting from the beam wave equation, which has a Schrödinger structure, on a hypercubic lattice of size $L$, with weak nonlinearity of strength $λ$, we show that the two point correlation function can be asymptotically expressed as the solution of the 3-wave kinetic equation at the kinetic limit under very general assumptions. The limit is taken in the physical order: first taking $L\to\infty$, then letting $λ\to0$. The initial condition is assumed out of equilibrium and the physical dimension is $d= 3$. The combination of previous works on the derivations of the wave kinetic equations from the lattice ZK equation [42,74] (with stochasticity, due to the singularity of the lattice ZK dispersion relation) and the present work (without stochasticity, thanks to the Schrödinger structure) give a detailed and general picture of how sensitive the kinetic descriptions are with respect to the regularities of the dispersion relations in the physical order of taking the limits. In terms of technical novelties, the current paper introduces new tools that produce the first obtained control of the Duhamel expansions when the number of collisions is of order $\mathcal{O}(λ^{-γ})$ (with $γ>0$) in the nonlinear context. To the best of our knowledge, such a strong control is a major step in understanding the long time dynamics of weakly nonlinear interactions of waves, and until now it could only be achieved in the linear context (see the work of Erdos-Salmhofer-Yau [34]). △ Less

Submitted 26 January, 2023; v1 submitted 19 January, 2023; originally announced January 2023.

Comments: Some of the proofs need some revisions. New version will be posted later

A Generalized Muirhead Inequality and Symmetric Sums of Nonnegative Circuits

Authors: Janin Heuer, Ngoc Mai Tran, Timo de Wolff

Abstract: Circuit polynomials are a certificate of nonnegativity for real polynomials, which can be derived via a generalization of the classical inequality of arithmetic and geometric means. In this article, we show that similarly nonnegativity of symmetric real polynomials can be certified via a generalization of the classical Muirhead inequality. Moreover, we show that a nonnegative symmetric polynomial… ▽ More Circuit polynomials are a certificate of nonnegativity for real polynomials, which can be derived via a generalization of the classical inequality of arithmetic and geometric means. In this article, we show that similarly nonnegativity of symmetric real polynomials can be certified via a generalization of the classical Muirhead inequality. Moreover, we show that a nonnegative symmetric polynomial admits a decomposition into sums of nonnegative circuit polynomials if and only if it satisfies said generalized Muirhead condition. The latter re-proves a result by Moustrou, Naumann, Riener, Theobald, and Verdure for the case of the symmetric group in a shortened and more elementary way. △ Less

Submitted 14 November, 2022; originally announced November 2022.

Comments: 16 pages

MSC Class: 12D15; 13J30; 14P99; 20B30; 26D15; 90C23

On the wave turbulence theory for a stochastic KdV type equation -- Generalization for the inhomogeneous kinetic limit

Authors: Amirali Hannani, Matthew Rosenzweig, Gigliola Staffilani, Minh-Binh Tran

Abstract: Starting from a stochastic Zakharov-Kuznetsov (ZK) equation on a lattice, the previous work [ST21] by the last two authors gave a derivation of the homogeneous 3-wave kinetic equation at the kinetic limit under very general assumptions: the initial condition is out of equilibrium, the dimension $d\ge 2$, the smallness of the nonlinearity $λ$ is allowed to be independent of the size of the lattice,… ▽ More Starting from a stochastic Zakharov-Kuznetsov (ZK) equation on a lattice, the previous work [ST21] by the last two authors gave a derivation of the homogeneous 3-wave kinetic equation at the kinetic limit under very general assumptions: the initial condition is out of equilibrium, the dimension $d\ge 2$, the smallness of the nonlinearity $λ$ is allowed to be independent of the size of the lattice, the weak noise is chosen not to compete with the weak nonlinearity and not to inject energy into the equation. In the present work, we build on the framework of [ST21], following the formal derivation of Spohn [Spo06] and inspired by previous work [HO21] of the first author and Olla, so that the inhomogeneous 3-wave kinetic equation can also be obtained at the kinetic limit under analogous assumptions. Similar to the homogeneous case -- and unlike the cubic nonlinear Schrödinger equation -- the inhomogeneous kinetic description of the deterministic lattice ZK equation is unlikely to happen due to the vanishing of the dispersion relation on a certain singular manifold on which not only $3$-wave interactions but also all $n$-wave interactions ($n\ge3$) are allowed to happen, a phenomenon first observed by Lukkarinen [Luk07]. To the best of our knowledge, our work provides the first rigorous derivation of a nonlinear inhomogeneous wave kinetic equation in the kinetic limit. △ Less

Submitted 8 November, 2022; v1 submitted 31 October, 2022; originally announced October 2022.

Comments: arXiv admin note: substantial text overlap with arXiv:2106.09819

A Deep Learning Approximation of Non-Stationary Solutions to Wave Kinetic Equations

Authors: Steven Walton, Minh-Binh Tran, Alain Bensoussan

Abstract: We present a deep learning approximation, stochastic optimization based, method for wave kinetic equations. To build confidence in our approach, we apply the method to a Smoluchowski coagulation equation with multiplicative kernel for which an analytic solution exists. Our deep learning approach is then used to approximate the non-stationary solution to a 3-wave kinetic equation corresponding to a… ▽ More We present a deep learning approximation, stochastic optimization based, method for wave kinetic equations. To build confidence in our approach, we apply the method to a Smoluchowski coagulation equation with multiplicative kernel for which an analytic solution exists. Our deep learning approach is then used to approximate the non-stationary solution to a 3-wave kinetic equation corresponding to acoustic wave systems. To validate the neural network approximation, we compare the decay rate of the total energy with previously obtained theoretical results. A finite volume solution is presented and compared with the present method. △ Less

Submitted 25 September, 2022; originally announced September 2022.

The tropical geometry of causal inference for extremes

Authors: Ngoc M Tran

Abstract: Extreme value statistics is the max analogue of classical statistics, while tropical geometry is the max analogue of classical geometry. In this paper, we review recent work where insights from tropical geometry were used to develop new, efficient learning algorithms with leading performance on benchmark datasets in extreme value statistics. We give intuition, backed by performances on benchmark d… ▽ More Extreme value statistics is the max analogue of classical statistics, while tropical geometry is the max analogue of classical geometry. In this paper, we review recent work where insights from tropical geometry were used to develop new, efficient learning algorithms with leading performance on benchmark datasets in extreme value statistics. We give intuition, backed by performances on benchmark datasets, for why and when causal inference for extremes should be employed over classical methods. Finally, we list some open problems at the intersection of causal inference, tropical geometry and deep learning. △ Less

Submitted 20 July, 2022; originally announced July 2022.

Comments: 19 pages, 5 figures

SHREC 2022: Fitting and recognition of simple geometric primitives on point clouds

Authors: Chiara Romanengo, Andrea Raffo, Silvia Biasotti, Bianca Falcidieno, Vlassis Fotis, Ioannis Romanelis, Eleftheria Psatha, Konstantinos Moustakas, Ivan Sipiran, Quang-Thuc Nguyen, Chi-Bien Chu, Khoi-Nguyen Nguyen-Ngoc, Dinh-Khoi Vo, Tuan-An To, Nham-Tan Nguyen, Nhat-Quynh Le-Pham, Hai-Dang Nguyen, Minh-Triet Tran, Yifan Qie, Nabil Anwer

Abstract: This paper presents the methods that have participated in the SHREC 2022 track on the fitting and recognition of simple geometric primitives on point clouds. As simple primitives we mean the classical surface primitives derived from constructive solid geometry, i.e., planes, spheres, cylinders, cones and tori. The aim of the track is to evaluate the quality of automatic algorithms for fitting and… ▽ More This paper presents the methods that have participated in the SHREC 2022 track on the fitting and recognition of simple geometric primitives on point clouds. As simple primitives we mean the classical surface primitives derived from constructive solid geometry, i.e., planes, spheres, cylinders, cones and tori. The aim of the track is to evaluate the quality of automatic algorithms for fitting and recognising geometric primitives on point clouds. Specifically, the goal is to identify, for each point cloud, its primitive type and some geometric descriptors. For this purpose, we created a synthetic dataset, divided into a training set and a test set, containing segments perturbed with different kinds of point cloud artifacts. Among the six participants to this track, two are based on direct methods, while four are either fully based on deep learning or combine direct and neural approaches. The performance of the methods is evaluated using various classification and approximation measures. △ Less

Submitted 7 July, 2022; v1 submitted 15 June, 2022; originally announced June 2022.

MSC Class: 68U05; 68U07; 65D18; 65D17 ACM Class: G.1.2; I.3.5; I.5.4

Journal ref: Computers & Graphics 107 (2022) 32-49

Minimal Representations of Tropical Rational Functions

Authors: Ngoc Mai Tran, Jidong Wang

Abstract: This paper studies the following question: given a piecewise-linear function, find its minimal algebraic representation as a tropical rational signomial. We put forward two different notions of minimality, one based on monomial length, the other based on factorization length. We show that in dimension one, both notions coincide, but this is not true in dimensions two or more. We prove uniqueness o… ▽ More This paper studies the following question: given a piecewise-linear function, find its minimal algebraic representation as a tropical rational signomial. We put forward two different notions of minimality, one based on monomial length, the other based on factorization length. We show that in dimension one, both notions coincide, but this is not true in dimensions two or more. We prove uniqueness of the minimal representation for dimension one and certain subclasses of piecewise-linear functions in dimension two. As a proof step, we obtain counting formulas and lower bounds for the number of regions in an arrangement of tropical hypersurfaces, giving a small extension for a result by Montúfar, Ren and Zhang. As an equivalent formulation, it gives a lower bound on the number of vertices in a regular mixed subdivision of a Minkowski sum, giving a small extension for Adiprasito's Lower Bound Theorem for Minkowski sums. △ Less

Submitted 16 January, 2024; v1 submitted 11 May, 2022; originally announced May 2022.

Comments: 27 pages, 15 figures

MSC Class: 14T15; 14T90; 52B05

Journal ref: Alg. Stat. 15 (2024) 27-59

Interpolative fusions II: Preservation results

Authors: Alex Kruckman, Minh Chieu Tran, Erik Walsberg

Abstract: We study interpolative fusion, a method of combining theories $T_1$ and $T_2$ in distinct languages in a "generic" way over a common reduct $T_\cap$, to obtain a theory $T_\cup^*$. When each $T_i$ is model-complete, $T_\cup^*$ is the model companion of the union $T_1\cup T_2$. Our goal is to prove preservation results, i.e., to find sufficient conditions under which model-theoretic properties of… ▽ More We study interpolative fusion, a method of combining theories $T_1$ and $T_2$ in distinct languages in a "generic" way over a common reduct $T_\cap$, to obtain a theory $T_\cup^*$. When each $T_i$ is model-complete, $T_\cup^*$ is the model companion of the union $T_1\cup T_2$. Our goal is to prove preservation results, i.e., to find sufficient conditions under which model-theoretic properties of $T_1$ and $T_2$ are inherited by $T_\cup^*$. We first prove preservation results for quantifier elimination, model-completeness, and related properties. We then apply these tools to show that, under mild hypotheses, including stability of $T_\cap$, the property $\mathrm{NSOP}_1$ is preserved. We also show that simplicity is preserved under stronger hypotheses on algebraic closure in $T_1$ and $T_2$. This generalizes many previous results; for example, simplicity of $\mathrm{ACFA}$ and the random $n$-hypergraph are both non-obvious corollaries. We also address preservation of stability, $\mathrm{NIP}$, and $\aleph_0$-categoricity, and we describe examples which witness that these results are sharp. △ Less

Submitted 10 January, 2022; originally announced January 2022.

Comments: Some parts of this paper originally appeared in the first version of the first interpolative fusion paper. arXiv admin note: text overlap with arXiv:1811.06108

A Numerical Scheme for Wave Turbulence: 3-Wave Kinetic Equations

Authors: Steven Walton, Minh-Binh Tran

Abstract: We introduce a finite volume scheme to solve a special case of isotropic 3-wave kinetic equations. We test our numerical solution against theoretical results concerning the long time behavior of the energy and observe that our solutions verify the energy cascade phenomenon. To our knowledge, this is the first numerical scheme that can capture the long time asymptotic behavior of solutions to those… ▽ More We introduce a finite volume scheme to solve a special case of isotropic 3-wave kinetic equations. We test our numerical solution against theoretical results concerning the long time behavior of the energy and observe that our solutions verify the energy cascade phenomenon. To our knowledge, this is the first numerical scheme that can capture the long time asymptotic behavior of solutions to those isotropic 3-wave kinetic equations, where the energy cascade can be observed. Our numerical energy cascade rates are in good agreement with previously obtained theoretical results. The finite volume scheme given here relies on a new identity, allowing one to reduce the number of terms needed in the collision operators. △ Less

Submitted 15 June, 2023; v1 submitted 19 October, 2021; originally announced October 2021.

Wave turbulence and collective behavior models for wave equations with short- and long-range interactions

Authors: Alejandro Aceves, Ricardo Alonso, Minh-Binh Tran

Abstract: In this work, we discuss a situation which could lead to both wave turbulence and collective behavior kinetic equations. The wave turbulence kinetic models appear in the kinetic limit when the wave equations have local differential operators. Viewing wave equations on the lattice as chains of anharmonic oscillators and replacing the local differential operators (short-range interactions) by non-lo… ▽ More In this work, we discuss a situation which could lead to both wave turbulence and collective behavior kinetic equations. The wave turbulence kinetic models appear in the kinetic limit when the wave equations have local differential operators. Viewing wave equations on the lattice as chains of anharmonic oscillators and replacing the local differential operators (short-range interactions) by non-local ones (long-range interactions), we arrive at a new Vlasov-type kinetic model in the mean field limit under the molecular chaos assumption reminiscent of models for collective behavior in which anharmonic oscillators replace individual particles. △ Less

Submitted 14 October, 2021; originally announced October 2021.

Minimax Rates for High-Dimensional Random Tessellation Forests

Authors: Eliza O'Reilly, Ngoc Mai Tran

Abstract: Random forests are a popular class of algorithms used for regression and classification. The algorithm introduced by Breiman in 2001 and many of its variants are ensembles of randomized decision trees built from axis-aligned partitions of the feature space. One such variant, called Mondrian forests, was proposed to handle the online setting and is the first class of random forests for which minima… ▽ More Random forests are a popular class of algorithms used for regression and classification. The algorithm introduced by Breiman in 2001 and many of its variants are ensembles of randomized decision trees built from axis-aligned partitions of the feature space. One such variant, called Mondrian forests, was proposed to handle the online setting and is the first class of random forests for which minimax rates were obtained in arbitrary dimension. However, the restriction to axis-aligned splits fails to capture dependencies between features, and random forests that use oblique splits have shown improved empirical performance for many tasks. In this work, we show that a large class of random forests with general split directions also achieve minimax optimal convergence rates in arbitrary dimension. This class includes STIT forests, a generalization of Mondrian forests to arbitrary split directions, as well as random forests derived from Poisson hyperplane tessellations. These are the first results showing that random forest variants with oblique splits can obtain minimax optimality in arbitrary dimension. Our proof technique relies on the novel application of the theory of stationary random tessellations in stochastic geometry to statistical learning theory. △ Less

Submitted 29 October, 2023; v1 submitted 22 September, 2021; originally announced September 2021.

Comments: 26 pages

MSC Class: 60D05; 62G07

Calderón-Zygmund-type estimates for singular quasilinear elliptic obstacle problems with measure data

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen, Phuoc-Nguyen Huynh

Abstract: We deal with a global Calderón-Zygmund type estimate for elliptic obstacle problems of $p$-Laplacian type with measure data. For this paper, we focus on the singular case of growth exponent, i.e. $1<p \le 2-\frac{1}{n}$. In addition, the emphasis of this paper is in obtaining the Lorentz bounds for the gradient of solutions with the use of fractional maximal operators. We deal with a global Calderón-Zygmund type estimate for elliptic obstacle problems of $p$-Laplacian type with measure data. For this paper, we focus on the singular case of growth exponent, i.e. $1<p \le 2-\frac{1}{n}$. In addition, the emphasis of this paper is in obtaining the Lorentz bounds for the gradient of solutions with the use of fractional maximal operators. △ Less

Submitted 15 December, 2021; v1 submitted 2 September, 2021; originally announced September 2021.

Comments: 28 pages

On the wave turbulence theory: ergodicity for the elastic beam wave equation

Authors: Benno Rumpf, Avy Soffer, Minh-Binh Tran

Abstract: We analyse a 3-wave kinetic equation, derived from the elastic beam wave equation on the lattice. The ergodicity condition states that two distinct wavevectors are supposed to be connected by a finite number of collisions. In this work, we prove that the ergodicity condition is violated and the equation domain is broken into disconnected domains, called no-collision and collisional invariant regio… ▽ More We analyse a 3-wave kinetic equation, derived from the elastic beam wave equation on the lattice. The ergodicity condition states that two distinct wavevectors are supposed to be connected by a finite number of collisions. In this work, we prove that the ergodicity condition is violated and the equation domain is broken into disconnected domains, called no-collision and collisional invariant regions. If one starts with a general initial condition, whose energy is finite, then in the long-time limit, the solutions of the 3-wave kinetic equation remain unchanged on the no-collision region and relax to local equilibria on the disjoint collisional invariant regions. △ Less

Submitted 10 March, 2025; v1 submitted 30 August, 2021; originally announced August 2021.

Competitive equilibrium always exists for combinatorial auctions with graphical pricing schemes

Authors: Marie-Charlotte Brandenburg, Christian Haase, Ngoc Mai Tran

Abstract: We show that a competitive equilibrium always exists in combinatorial auctions with anonymous graphical valuations and pricing, using discrete geometry. This is an intuitive and easy-to-construct class of valuations that can model both complementarity and substitutes, and to our knowledge, it is the first class besides gross substitutes that have guaranteed competitive equilibrium. We prove throug… ▽ More We show that a competitive equilibrium always exists in combinatorial auctions with anonymous graphical valuations and pricing, using discrete geometry. This is an intuitive and easy-to-construct class of valuations that can model both complementarity and substitutes, and to our knowledge, it is the first class besides gross substitutes that have guaranteed competitive equilibrium. We prove through counter-examples that our result is tight, and we give explicit algorithms for constructive competitive pricing vectors. We also give extensions to multi-unit combinatorial auctions (also known as product-mix auctions). Combined with theorems on graphical valuations and pricing equilibrium of Candogan, Ozdagar and Parrilo, our results indicate that quadratic pricing is a highly practical method to run combinatorial auctions. △ Less

Submitted 5 November, 2021; v1 submitted 19 July, 2021; originally announced July 2021.

Comments: 24 pages, 5 figures

MSC Class: 91B26; 14T90; 52B20

Quantitative Marcinkiewicz's theorem and central limit theorems: applications to spin systems and point processes

Authors: Tien-Cuong Dinh, Subhroshekhar Ghosh, Hoang-Son Tran, Manh-Hung Tran

Abstract: The classical Marcinkiewicz theorem states that if an entire characteristic function $Ψ$ of a non-degenerate real-valued random variable $X$ is of the form $\exp(P(u))$ for some polynomial $P$, then $X$ has to be a Gaussian. In this work, we obtain a broad, quantitative extension of this framework in several directions, establish central limit theorems (CLTs) with explicit rates of convergence, an… ▽ More The classical Marcinkiewicz theorem states that if an entire characteristic function $Ψ$ of a non-degenerate real-valued random variable $X$ is of the form $\exp(P(u))$ for some polynomial $P$, then $X$ has to be a Gaussian. In this work, we obtain a broad, quantitative extension of this framework in several directions, establish central limit theorems (CLTs) with explicit rates of convergence, and demonstrate Gaussian fluctuations in continuous spin systems and general classes of point processes. In particular, we obtain quantitative decay estimates on the Kolmogorov-Smirnov distance between $X$ and a Gaussian under the condition that $Ψ$ does not vanish only on a bounded disk. This leads to quantitative CLTs applicable to very general and possibly strongly dependent random systems. In spite of the general applicability, our rates for the CLT match the classic Berry-Esseen bounds for independent sums up to a log factor. We implement this programme for two important classes of models in probability and statistical physics. First, we extend to the setting of continuous spins a popular paradigm for obtaining CLTs for discrete spin systems that is based on the theory of Lee-Yang zeros, focussing in particular on the XY model, Heisenberg ferromagnets and generalised Ising models. Secondly, we establish Gaussian fluctuations for linear statistics of so-called $α$-determinantal processes for $α\in \mathbb{R}$ (including the usual determinantal, Poisson and permanental processes) under very general conditions, including in particular higher dimensional settings where structural alternatives such as random matrix techniques are not available. Our applications demonstrate the significance of having to control the characteristic function only on a (small) disk, and lead to CLTs which, to the best of our knowledge, are not known in generality. △ Less

Submitted 25 July, 2021; v1 submitted 18 July, 2021; originally announced July 2021.

On the wave turbulence theory for a stochastic KdV type equation

Authors: Gigliola Staffilani, Minh-Binh Tran

Abstract: Starting from the stochastic Zakharov-Kuznetsov equation, a multidimensional KdV type equation on a hypercubic lattice, we provide a derivation of the 3-wave kinetic equation. We show that the two point correlation function can be asymptotically expressed as the solution of the 3-wave kinetic equation at the kinetic limit under very general assumptions: the initial condition is out of equilibrium,… ▽ More Starting from the stochastic Zakharov-Kuznetsov equation, a multidimensional KdV type equation on a hypercubic lattice, we provide a derivation of the 3-wave kinetic equation. We show that the two point correlation function can be asymptotically expressed as the solution of the 3-wave kinetic equation at the kinetic limit under very general assumptions: the initial condition is out of equilibrium, the dimension is $d\ge 2$, the smallness of the nonlinearity $λ$ is allowed to be independent of the size of the lattice, the weak noise is chosen not to compete with the weak nonlinearity and not to inject energy into the equation. Unlike the cubic nonlinear Schrödinger equation, for which such a general result is commonly expected without the noise, the kinetic description of the deterministic lattice ZK equation is unlikely to happen. One of the key reasons is that the dispersion relation of the lattice ZK equation leads to a singular manifold, on which not only $3$-wave interactions but also all $m$-wave interactions ($m\ge3$) are allowed to happen. This phenomenon has been first observed by Lukkarinen \cite{lukkarinen2007asymptotics} as a counterexample for which one of the main tools to derive kinetic equations from wave equations (the suppression of crossings) fails to hold true. To the best of our knowledge, the work provides the first rigorous derivation of nonlinear 3-wave kinetic equations. Also this is the first derivation for wave kinetic equations in the lattice setting and out-of-equilibrium. △ Less

Submitted 12 February, 2024; v1 submitted 17 June, 2021; originally announced June 2021.

A hybrid discrete-continuum approach to model hydro-mechanical behaviour of soil during desiccation

Authors: Khoa M. Tran, Ha H. Bui, Giang D. Nguyen

Abstract: Desiccation cracking in clayey soils occurs when they lose moisture, leading to an increase in their compressibility and hydraulic conductivity and hence significant reduction of soil strength. The prediction of desiccation cracking in soils is challenging due to the lack of insights into the complex coupled hydro-mechanical process at the grain scale. In this paper, a new hybrid discrete-continuu… ▽ More Desiccation cracking in clayey soils occurs when they lose moisture, leading to an increase in their compressibility and hydraulic conductivity and hence significant reduction of soil strength. The prediction of desiccation cracking in soils is challenging due to the lack of insights into the complex coupled hydro-mechanical process at the grain scale. In this paper, a new hybrid discrete-continuum numerical framework, capable of capturing hydro-mechanical behaviour of soil at both grain and macro scales, is proposed for predicting desiccation cracking in clayey soil. In this framework, a soil layer is represented by an assembly of DEM particles, each occupies an equivalent continuum space and carries physical properties governing unsaturated flow. These particles move freely in the computational space following the discrete element method (DEM), while their contact network and the continuum mixture theory are used to model the unsaturated flow. The dependence of particle-to-particle contact behaviour on water content is represented by a cohesive-frictional contact model, whose material properties are governed by the water content. In parallel with the theoretical development is a series of experiments on 3D soil desiccation cracking to determine essential properties and provide data for the validation of mechanical and physical behaviour. Very good agreement in both physical behaviour (e.g. evolution of water content) and mechanical behaviour (e.g. occurrence and development of cracks, and distribution of compressive and tensile strains) demonstrates that the proposed framework is capable of capturing the hydro-mechanical behaviour of soil during desiccation. The capability of the proposed framework facilitates numerical experiments for insights into the hydro-mechanical behaviour of unsaturated soils that have not been possible before. △ Less

Submitted 25 May, 2021; originally announced June 2021.

New oscillation classes and two weight bump conditions for commutators

Authors: David Cruz-Uribe, Kabe Moen, Quan Minh Tran

Abstract: In this paper we consider two weight bump conditions for higher order commutators. Given $b$ and a Calderón-Zygmund operator $T$, define the commutator $T^1_bf=[T,b]f= bTf-T(bf)$, and for $m\geq 2$ define the iterated commutator $T^m_b f = [b,T_b^{m-1}]f$. Traditionally, commutators are defined for functions $b\in BMO$, but we show that if we replace $BMO$ by an oscillation class first introduced… ▽ More In this paper we consider two weight bump conditions for higher order commutators. Given $b$ and a Calderón-Zygmund operator $T$, define the commutator $T^1_bf=[T,b]f= bTf-T(bf)$, and for $m\geq 2$ define the iterated commutator $T^m_b f = [b,T_b^{m-1}]f$. Traditionally, commutators are defined for functions $b\in BMO$, but we show that if we replace $BMO$ by an oscillation class first introduced by Pérez [31], we can give a range of sufficient conditions on a pair of weights $(u,v)$ for $T^m_b : L^p(v)\rightarrow L^p(u)$ to be bounded. Our results generalize work of the first two authors in [10], and more recent work by Lerner, et al. [28]. We also prove necessary conditions for the iterated commutators to be bounded, generalizing results of Isralowitz, et al.[20]. △ Less

Submitted 11 March, 2021; originally announced March 2021.

Lorentz gradient estimates for a class of elliptic p-Laplacian equations with a Schrödinger term

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen, Gia-Bao Nguyen

Abstract: We prove in this paper the global Lorentz estimate in term of fractional-maximal function for gradient of weak solutions to a class of p-Laplace elliptic equations containing a non-negative Schrödinger potential which belongs to reverse Hölder classes. In particular, this class of p-Laplace operator includes both degenerate and non-degenerate cases. The interesting idea is to use an efficient appr… ▽ More We prove in this paper the global Lorentz estimate in term of fractional-maximal function for gradient of weak solutions to a class of p-Laplace elliptic equations containing a non-negative Schrödinger potential which belongs to reverse Hölder classes. In particular, this class of p-Laplace operator includes both degenerate and non-degenerate cases. The interesting idea is to use an efficient approach based on the level-set inequality related to the distribution function in harmonic analysis. △ Less

Submitted 26 September, 2020; originally announced September 2020.

Comments: 19 pages

Discriminants of cubic curves and determinantal representations

Authors: Manh Hung Tran

Abstract: The discriminant of a smooth plane cubic curve over the complex numbers can be written as a product of theta functions. This provides an important connection between algebraic and analytic objects. In this paper, we perform a new approach to obtain this classical result by using determinantal representations. More precisely, one can represent a non-singular cubic form as the determinant of a matri… ▽ More The discriminant of a smooth plane cubic curve over the complex numbers can be written as a product of theta functions. This provides an important connection between algebraic and analytic objects. In this paper, we perform a new approach to obtain this classical result by using determinantal representations. More precisely, one can represent a non-singular cubic form as the determinant of a matrix whose elements are linear forms. Theta functions naturally appear in this representation and thus in the discriminant of the cubic. △ Less

Submitted 2 May, 2022; v1 submitted 14 September, 2020; originally announced September 2020.

Comments: 16 pages. To appear in The Ramanujan Journal. Portions of this work previously appeared as arXiv:1911.01350 which was split during refereeing

MSC Class: 11C20; 11G07; 14H42; 14H50; 14H55; 32A08

A global fractional Caccioppoli-type estimate for solutions to nonlinear elliptic problems with measure data

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen

Abstract: We prove a global fractional differentiability result via the fractional Caccioppoli-type estimate for solutions to nonlinear elliptic problems with measure data. This work is in fact inspired by the recent paper [B. Avelin, T. Kuusi, G. Mingione, {\em Nonlinear Calderón-Zygmund theory in the limiting case}, Arch. Rational. Mech. Anal. {\bf 227}(2018), 663--714], that was devoted to the local frac… ▽ More We prove a global fractional differentiability result via the fractional Caccioppoli-type estimate for solutions to nonlinear elliptic problems with measure data. This work is in fact inspired by the recent paper [B. Avelin, T. Kuusi, G. Mingione, {\em Nonlinear Calderón-Zygmund theory in the limiting case}, Arch. Rational. Mech. Anal. {\bf 227}(2018), 663--714], that was devoted to the local fractional regularity for the solutions to nonlinear elliptic equations with right-hand side measure, of type $-\mathrm{div}\, \mathcal{A}(\nabla u) = μ$ in the limiting case. Being a contribution to recent results of identifying function classes that solutions to such problems could be defined, our aim in this work is to establish a global regularity result in a setting of weighted fractional Sobolev spaces, where the weights are powers of the distance function to the boundary of the smooth domains. △ Less

Submitted 6 September, 2020; originally announced September 2020.

Comments: 13 pages

Machine Learning and Control Theory

Authors: Alain Bensoussan, Yiqun Li, Dinh Phan Cao Nguyen, Minh-Binh Tran, Sheung Chi Phillip Yam, Xiang Zhou

Abstract: We survey in this article the connections between Machine Learning and Control Theory. Control Theory provide useful concepts and tools for Machine Learning. Conversely Machine Learning can be used to solve large control problems. In the first part of the paper, we develop the connections between reinforcement learning and Markov Decision Processes, which are discrete time control problems. In the… ▽ More We survey in this article the connections between Machine Learning and Control Theory. Control Theory provide useful concepts and tools for Machine Learning. Conversely Machine Learning can be used to solve large control problems. In the first part of the paper, we develop the connections between reinforcement learning and Markov Decision Processes, which are discrete time control problems. In the second part, we review the concept of supervised learning and the relation with static optimization. Deep learning which extends supervised learning, can be viewed as a control problem. In the third part, we present the links between stochastic gradient descent and mean-field theory. Conversely, in the fourth and fifth parts, we review machine learning approaches to stochastic control problems, and focus on the deterministic case, to explain, more easily, the numerical algorithms. △ Less

Submitted 9 June, 2020; originally announced June 2020.

Weighted distribution approach to gradient estimates for quasilinear elliptic double-obstacle problems in Orlicz spaces

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: We construct an efficient approach to deal with the global regularity estimates for a class of elliptic double-obstacle problems in Lorentz and Orlicz spaces. The motivation of this paper comes from the study on an abstract result in the viewpoint of the fractional maximal distributions and this work also extends some regularity results proved in \cite{PN_dist} by using the weighted fractional max… ▽ More We construct an efficient approach to deal with the global regularity estimates for a class of elliptic double-obstacle problems in Lorentz and Orlicz spaces. The motivation of this paper comes from the study on an abstract result in the viewpoint of the fractional maximal distributions and this work also extends some regularity results proved in \cite{PN_dist} by using the weighted fractional maximal distributions (WFMDs). We further investigate a pointwise estimates of the gradient of weak solutions via fractional maximal operators and Riesz potential of data. Moreover, in the setting of the paper, we are led to the study of problems with nonlinearity is supposed to be partially weak BMO condition (is measurable in one fixed variable and only satisfies locally small-BMO seminorms in the remaining variables). △ Less

Submitted 4 June, 2020; originally announced June 2020.

Identification of linear dynamical systems and machine learning

Authors: Alain Bensoussan, Fatih Gelir, Viswanath Ramakrishna, Minh-Binh Tran

Abstract: The topic of identification of dynamic systems, has been at the core of modern control , following the fundamental works of Kalman. Realization Theory has been one of the major outcomes in this domain, with the possibility of identifying a dynamic system from an input-output relationship. The recent development of machine learning concepts has rejuvanated interest for identification. In this paper… ▽ More The topic of identification of dynamic systems, has been at the core of modern control , following the fundamental works of Kalman. Realization Theory has been one of the major outcomes in this domain, with the possibility of identifying a dynamic system from an input-output relationship. The recent development of machine learning concepts has rejuvanated interest for identification. In this paper, we review briefly the results of realization theory, and develop some methods inspired by Machine Learning concepts. △ Less

Submitted 15 May, 2020; originally announced May 2020.

Lorentz estimates for quasi-linear elliptic double obstacle problems involving a Schrödinger term

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: Our goal in this article is to study the global Lorentz estimates for gradient of weak solutions to $p$-Laplace double obstacle problems involving the Schrödinger term: $-Δ_p u + \mathbb{V}|u|^{p-2}u$ with bound constraints $ψ_1 \le u \le ψ_2$ in non-smooth domains. This problem has its own interest in mathematics, engineering, physics and other branches of science. Our approach makes a novel conn… ▽ More Our goal in this article is to study the global Lorentz estimates for gradient of weak solutions to $p$-Laplace double obstacle problems involving the Schrödinger term: $-Δ_p u + \mathbb{V}|u|^{p-2}u$ with bound constraints $ψ_1 \le u \le ψ_2$ in non-smooth domains. This problem has its own interest in mathematics, engineering, physics and other branches of science. Our approach makes a novel connection between the study of Calderón-Zygmund theory for nonlinear Schrödinger type equations and variational inequalities for double obstacle problems. △ Less

Submitted 7 May, 2020; originally announced May 2020.

Comments: 20 pages

Level-set inequalities on fractional maximal distribution functions and applications to regularity theory

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: The aim of this paper is to establish an abstract theory based on the so-called fractional-maximal distribution functions (FMDs). From the rough ideas introduced in~\cite{AM2007}, we develop and prove some abstract results related to the level-set inequalities and norm-comparisons by using the language of such FMDs. Particularly interesting is the applicability of our approach that has been shown… ▽ More The aim of this paper is to establish an abstract theory based on the so-called fractional-maximal distribution functions (FMDs). From the rough ideas introduced in~\cite{AM2007}, we develop and prove some abstract results related to the level-set inequalities and norm-comparisons by using the language of such FMDs. Particularly interesting is the applicability of our approach that has been shown in regularity and Calderón-Zygmund type estimates. In this paper, due to our research experience, we will establish global regularity estimates for two types of general quasilinear problems (problems with divergence form and double obstacles), via fractional-maximal operators and FMDs. The range of applications of these abstract results is large. Apart from these two examples of the regularity theory for elliptic equations discussed, it is also promising to indicate further possible applications of our approach for other special topics. △ Less

Submitted 14 April, 2020; originally announced April 2020.

Comments: 41 pages

Lorentz improving estimates for the $p$-Laplace equations with mixed data

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: The aim of this paper is to develop the regularity theory for a weak solution to a class of quasilinear nonhomogeneous elliptic equations, whose prototype is the following mixed Dirichlet $p$-Laplace equation of type \begin{align*} \begin{cases} \mathrm{div}(|\nabla u|^{p-2}\nabla u) &= f+ \ \mathrm{div}(|\mathbf{F}|^{p-2}\mathbf{F}) \qquad \text{in} \ \ Ω, \\ \hspace{1.2cm} u &=\ g \hspace{3.1cm}… ▽ More The aim of this paper is to develop the regularity theory for a weak solution to a class of quasilinear nonhomogeneous elliptic equations, whose prototype is the following mixed Dirichlet $p$-Laplace equation of type \begin{align*} \begin{cases} \mathrm{div}(|\nabla u|^{p-2}\nabla u) &= f+ \ \mathrm{div}(|\mathbf{F}|^{p-2}\mathbf{F}) \qquad \text{in} \ \ Ω, \\ \hspace{1.2cm} u &=\ g \hspace{3.1cm} \text{on} \ \ \partial Ω, \end{cases} \end{align*} in Lorentz space, with given data $\mathbf{F} \in L^p(Ω;\mathbb{R}^n)$, $f \in L^{\frac{p}{p-1}}(Ω)$, $g \in W^{1,p}(Ω)$ for $p>1$ and $Ω\subset \mathbb{R}^n$ ($n \ge 2$) satisfying a Reifenberg flat domain condition or a $p$-capacity uniform thickness condition, which are considered in several recent papers. To better specify our result, the proofs of regularity estimates involve fractional maximal operators and valid for a more general class of quasilinear nonhomogeneous elliptic equations with mixed data. This paper not only deals with the Lorentz estimates for a class of more general problems with mixed data but also improves the good-$λ$ approach technique proposed in our preceding works~\cite{MPT2018,PNCCM,PNJDE,PNCRM}, to achieve the global Lorentz regularity estimates for gradient of weak solutions in terms of fractional maximal operators. △ Less

Submitted 10 March, 2020; v1 submitted 10 March, 2020; originally announced March 2020.

Global Lorentz estimates for non-uniformly nonlinear elliptic equations via fractional maximal operators

Authors: Thanh-Nhan Nguyen, Minh-Phuong Tran

Abstract: This paper is a contribution to the study of regularity theory for nonlinear elliptic equations. The aim of this paper is to establish some global estimates for non-uniformly elliptic in divergence form as follows \begin{align*} -\mathrm{div}(|\nabla u|^{p-2}\nabla u + a(x)|\nabla u|^{q-2}\nabla u) = - \mathrm{div}(|\mathbf{F}|^{p-2}\mathbf{F} + a(x)|\mathbf{F}|^{q-2}\mathbf{F}), \end{align*} that… ▽ More This paper is a contribution to the study of regularity theory for nonlinear elliptic equations. The aim of this paper is to establish some global estimates for non-uniformly elliptic in divergence form as follows \begin{align*} -\mathrm{div}(|\nabla u|^{p-2}\nabla u + a(x)|\nabla u|^{q-2}\nabla u) = - \mathrm{div}(|\mathbf{F}|^{p-2}\mathbf{F} + a(x)|\mathbf{F}|^{q-2}\mathbf{F}), \end{align*} that arises from double phase functional problems. In particular, the main results provide the regularity estimates for the distributional solutions in terms of maximal and fractional maximal operators. This work extends that of \cite{CoMin2016,Byun2017Cava} by dealing with the global estimates in Lorentz spaces. This work also extends our recent result in \cite{PNJDE}, which is devoted to the new estimates of divergence elliptic equations using cut-off fractional maximal operators. For future research, the approach developed in this paper allows to attain global estimates of distributional solutions to non-uniformly nonlinear elliptic equations in the framework of other spaces. △ Less

Submitted 1 February, 2020; originally announced February 2020.

A Theory of Trotter Error

Authors: Andrew M. Childs, Yuan Su, Minh C. Tran, Nathan Wiebe, Shuchen Zhu

Abstract: The Lie-Trotter formula, together with its higher-order generalizations, provides a direct approach to decomposing the exponential of a sum of operators. Despite significant effort, the error scaling of such product formulas remains poorly understood. We develop a theory of Trotter error that overcomes the limitations of prior approaches based on truncating the Baker-Campbell-Hausdorff expansion.… ▽ More The Lie-Trotter formula, together with its higher-order generalizations, provides a direct approach to decomposing the exponential of a sum of operators. Despite significant effort, the error scaling of such product formulas remains poorly understood. We develop a theory of Trotter error that overcomes the limitations of prior approaches based on truncating the Baker-Campbell-Hausdorff expansion. Our analysis directly exploits the commutativity of operator summands, producing tighter error bounds for both real- and imaginary-time evolutions. Whereas previous work achieves similar goals for systems with geometric locality or Lie-algebraic structure, our approach holds in general. We give a host of improved algorithms for digital quantum simulation and quantum Monte Carlo methods, including simulations of second-quantized plane-wave electronic structure, $k$-local Hamiltonians, rapidly decaying power-law interactions, clustered Hamiltonians, the transverse field Ising model, and quantum ferromagnets, nearly matching or even outperforming the best previous results. We obtain further speedups using the fact that product formulas can preserve the locality of the simulated system. Specifically, we show that local observables can be simulated with complexity independent of the system size for power-law interacting systems, which implies a Lieb-Robinson bound as a byproduct. Our analysis reproduces known tight bounds for first- and second-order formulas. Our higher-order bound overestimates the complexity of simulating a one-dimensional Heisenberg model with an even-odd ordering of terms by only a factor of $5$, and is close to tight for power-law interactions and other orderings of terms. This suggests that our theory can accurately characterize Trotter error in terms of both asymptotic scaling and constant prefactor. △ Less

Submitted 3 February, 2021; v1 submitted 18 December, 2019; originally announced December 2019.

Comments: 82 pages, 5 figures. Enhanced version of the article published in Physical Review X at http://journals.aps.org/prx/abstract/10.1103/PhysRevX.11.011020

Journal ref: Phys. Rev. X 11, 011020 (2021)

Invariants of models of genus one curves and modular forms

Authors: Manh Hung Tran

Abstract: An invariant of a model of genus one curve is a polynomial in the coefficients of the model that is stable under certain linear transformations. The classical example of an invariant is the discriminant, which characterizes the singularity of models. The ring of invariants of genus one models over a field is generated by two elements. Fisher normalized these invariants for models of degree n=2,3,4… ▽ More An invariant of a model of genus one curve is a polynomial in the coefficients of the model that is stable under certain linear transformations. The classical example of an invariant is the discriminant, which characterizes the singularity of models. The ring of invariants of genus one models over a field is generated by two elements. Fisher normalized these invariants for models of degree n=2,3,4 in such a way that these invariants are moreover defined over the integers. We provide an alternative way to express these normalized invariants using modular forms. This method relies on a direct computation for the discriminants based on their own geometric properties. △ Less

Submitted 2 September, 2020; v1 submitted 4 November, 2019; originally announced November 2019.

Comments: 18 pages. Updated version consisting of the first 4 sections from the original one, introduction rewritten

MSC Class: 11F11; 11G05; 11Y40; 13A50; 14H50; 14Q05

An application of global gradient estimates in Lorentz-Morrey spaces: The existence of stationary solution to degenerate diffusive Hamilton-Jacobi equations

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen

Abstract: In historical mathematics and physics, the Kardar-Parisi-Zhang equation or a quasilinear stationary version of a time-dependent viscous Hamilton-Jacobi equation in growing interface and universality classes, is also known by the different name as the quasilinear Riccati type equation. The existence of solutions to this type of equation under some assumptions and requirements, still remains an inte… ▽ More In historical mathematics and physics, the Kardar-Parisi-Zhang equation or a quasilinear stationary version of a time-dependent viscous Hamilton-Jacobi equation in growing interface and universality classes, is also known by the different name as the quasilinear Riccati type equation. The existence of solutions to this type of equation under some assumptions and requirements, still remains an interesting open problem at the moment. In our previous studies \cite{MP2018, MPT2019}, we obtained the global bounds and gradient estimates for quasilinear elliptic equations with measure data. There have been many applications are discussed related to these works, and main goal of this paper is to obtain the existence of a renormalized solution to the quasilinear stationary solution to the degenerate diffusive Hamilton-Jacobi equation with the finite measure data in Lorentz-Morrey spaces. △ Less

Submitted 10 October, 2019; originally announced October 2019.

Global gradient estimates for very singular quasilinear elliptic equations with measure data

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen

Abstract: This paper continues the development of regularity results for quasilinear measure data problems \begin{align*} \begin{cases} -\mathrm{div}(A(x,\nabla u)) &= μ\quad \text{in} \ \ Ω, \\ \quad \quad \qquad u &=0 \quad \text{on} \ \ \partial Ω, \end{cases} \end{align*} in Lorentz and Lorentz-Morrey spaces, where $Ω\subset \mathbb{R}^n$ ($n \ge 2$), $μ$ is a finite Radon measure on $Ω$, and $A$ is a m… ▽ More This paper continues the development of regularity results for quasilinear measure data problems \begin{align*} \begin{cases} -\mathrm{div}(A(x,\nabla u)) &= μ\quad \text{in} \ \ Ω, \\ \quad \quad \qquad u &=0 \quad \text{on} \ \ \partial Ω, \end{cases} \end{align*} in Lorentz and Lorentz-Morrey spaces, where $Ω\subset \mathbb{R}^n$ ($n \ge 2$), $μ$ is a finite Radon measure on $Ω$, and $A$ is a monotone Carathéodory vector valued operator acting between $W^{1,p}_0(Ω)$ and its dual $W^{-1,p'}(Ω)$. It emphasizes that this paper studies the `very singular' case $1<p \le \frac{3n-2}{2n-1}$ and the problem is considered under the weak assumption, where the $p$-capacity uniform thickness condition is imposed on the complement of domain $Ω$. There are two main results obtained in our study pertaining to the global gradient estimates of solutions in Lorentz and Lorentz-Morrey spaces involving the use of maximal and fractional maximal operators. The idea for writing this working paper comes directly from the recent results by others in the same research topic, where global estimates for gradient of solutions for the `very singular' case still remains a challenge, specifically related to Lorentz and Lorentz-Morrey spaces. △ Less

Submitted 5 April, 2021; v1 submitted 16 September, 2019; originally announced September 2019.

Comments: 33 pages

Statistics and tropicalization of local field Gaussian measures

Authors: Yassine El Maazouz, Ngoc Mai Tran

Abstract: This paper aims to lay the foundations for statistics over local fields, such as the field of $p$-adic numbers. Over such fields, we give characterizations for maximum likelihood estimation and conditional independence for multivariate Gaussian distributions. We also give a bijection between the tropicalization of such Gaussian measures in dimension 2 and supermodular functions on the 2-dimensiona… ▽ More This paper aims to lay the foundations for statistics over local fields, such as the field of $p$-adic numbers. Over such fields, we give characterizations for maximum likelihood estimation and conditional independence for multivariate Gaussian distributions. We also give a bijection between the tropicalization of such Gaussian measures in dimension 2 and supermodular functions on the 2-dimensional discrete cube. Finally, we introduce the Bruhat-Tits building as a parameter space for Gaussian distributions and discuss their connections to conditional independence statements as an open problem. △ Less

Submitted 16 August, 2021; v1 submitted 2 September, 2019; originally announced September 2019.

Comments: 15 pages, 4 figures

MSC Class: 62H05; 60E05; 12J25; 14T90

Generalized potential games

Authors: M. H. Duong, T. H. Dang-Ha, Q. B. Tang, H. M. Tran

Abstract: In this paper, we introduce a notion of generalized potential games that is inspired by a newly developed theory on generalized gradient flows. More precisely, a game is called generalized potential if the simultaneous gradient of the loss functions is a nonlinear function of the gradient of a potential function. Applications include a class of games arising from chemical reaction networks with de… ▽ More In this paper, we introduce a notion of generalized potential games that is inspired by a newly developed theory on generalized gradient flows. More precisely, a game is called generalized potential if the simultaneous gradient of the loss functions is a nonlinear function of the gradient of a potential function. Applications include a class of games arising from chemical reaction networks with detailed balance condition. For this class of games, we prove an explicit exponential convergence to equilibrium for evolution of a single reversible reaction. Moreover, numerical investigations are performed to calculate the equilibrium state of some reversible chemical reactions which give rise to generalized potential games. △ Less

Submitted 17 August, 2019; originally announced August 2019.

Comments: 23 pages, 6 figures. Comments are welcome

Gradient estimates via Riesz potentials and fractional maximal operators for quasilinear elliptic equations with applications

Authors: Minh-Phuong Tran, Thanh-Nhan Nguyen

Abstract: In this paper, the aim of our work is to establish global weighted gradient estimates via fractional maximal functions and the point-wise regularity estimates of Dirichlet problem for divergence elliptic equations of the type \begin{align*} \mathrm{div}(A(x,\nabla u)) = \mathrm{div}(f) \ \text{in} \ Ω, \mbox{ and } \ u = g \ \text{on} \ \partial Ω, \end{align*} that related to Riesz potentials. He… ▽ More In this paper, the aim of our work is to establish global weighted gradient estimates via fractional maximal functions and the point-wise regularity estimates of Dirichlet problem for divergence elliptic equations of the type \begin{align*} \mathrm{div}(A(x,\nabla u)) = \mathrm{div}(f) \ \text{in} \ Ω, \mbox{ and } \ u = g \ \text{on} \ \partial Ω, \end{align*} that related to Riesz potentials. Here, in our setting, $Ω\subset \mathbb{R}^n$, $n \ge 2$ is a bounded Reifenberg flat domain (that its boundary is sufficiently flat in sense of Reifenberg) and the small-BMO condition (small bounded mean oscillations) is assumed on the nonlinearity $A$. Further, the emphasis of the paper is the existence of weak solution to a class of quasilinear elliptic equations containing Riesz potential of the gradient term, as an application of the global point-wise bound. And regarding this study, we also analyze the necessary and sufficient conditions that guarantee the existence of solution to such nonlinear elliptic problems. △ Less

Submitted 19 July, 2021; v1 submitted 2 July, 2019; originally announced July 2019.

Comments: 30 pages

On the Wave Turbulence Theory for the Nonlinear Schrödinger Equation with Random Potentials

Authors: Sergey Nazarenko, Avy Soffer, Minh-Binh Tran

Abstract: We derive a new kinetic and a porous medium equations from the nonlinear Schrödinger equation with random potentials. The kinetic equation has a very similar form with the 4-wave turbulence kinetic equation in the wave turbulence theory. Moreover, we construct a class of self-similar solutions for the porous medium equation. These solutions spread infinitely as time goes to infinity and this fact… ▽ More We derive a new kinetic and a porous medium equations from the nonlinear Schrödinger equation with random potentials. The kinetic equation has a very similar form with the 4-wave turbulence kinetic equation in the wave turbulence theory. Moreover, we construct a class of self-similar solutions for the porous medium equation. These solutions spread infinitely as time goes to infinity and this fact answers the 'weak turbulence' question for the nonlinear Schrödinger equation with random potentials positively. We also derive Ohm's law for the porous medium equation. △ Less

Submitted 14 May, 2019; originally announced May 2019.