-
Empirical Tests of Optimization Assumptions in Deep Learning
Authors:
Hoang Tran,
Qinzi Zhang,
Ashok Cutkosky
Abstract:
There is a significant gap between our theoretical understanding of optimization algorithms used in deep learning and their practical performance. Theoretical development usually focuses on proving convergence guarantees under a variety of different assumptions, which are themselves often chosen based on a rough combination of intuitive match to practice and analytical convenience. The theory/prac…
▽ More
There is a significant gap between our theoretical understanding of optimization algorithms used in deep learning and their practical performance. Theoretical development usually focuses on proving convergence guarantees under a variety of different assumptions, which are themselves often chosen based on a rough combination of intuitive match to practice and analytical convenience. The theory/practice gap may then arise because of the failure to prove a theorem under such assumptions, or because the assumptions do not reflect reality. In this paper, we carefully measure the degree to which these assumptions are capable of explaining modern optimization algorithms by develo** new empirical metrics that closely track the key quantities that must be controlled in theoretical analysis. All of our tested assumptions (including typical modern assumptions based on bounds on the Hessian) fail to reliably capture optimization performance. This highlights a need for new empirical verification of analytical assumptions used in theoretical analysis.
△ Less
Submitted 1 July, 2024;
originally announced July 2024.
-
Private Zeroth-Order Nonsmooth Nonconvex Optimization
Authors:
Qinzi Zhang,
Hoang Tran,
Ashok Cutkosky
Abstract:
We introduce a new zeroth-order algorithm for private stochastic optimization on nonconvex and nonsmooth objectives. Given a dataset of size $M$, our algorithm ensures $(α,αρ^2/2)$-Rényi differential privacy and finds a $(δ,ε)$-stationary point so long as $M=\tildeΩ\left(\frac{d}{δε^3} + \frac{d^{3/2}}{ρδε^2}\right)$. This matches the optimal complexity of its non-private zeroth-order analog. Nota…
▽ More
We introduce a new zeroth-order algorithm for private stochastic optimization on nonconvex and nonsmooth objectives. Given a dataset of size $M$, our algorithm ensures $(α,αρ^2/2)$-Rényi differential privacy and finds a $(δ,ε)$-stationary point so long as $M=\tildeΩ\left(\frac{d}{δε^3} + \frac{d^{3/2}}{ρδε^2}\right)$. This matches the optimal complexity of its non-private zeroth-order analog. Notably, although the objective is not smooth, we have privacy ``for free'' whenever $ρ\ge \sqrt{d}ε$.
△ Less
Submitted 27 June, 2024;
originally announced June 2024.
-
Analytic bounds on late-time axion-scalar cosmologies
Authors:
Gary Shiu,
Flavio Tonioni,
Hung V. Tran
Abstract:
The cosmological dynamics of multiple scalar/pseudoscalar fields are difficult to solve, especially when the field-space metric is curved. This presents a challenge in determining whether a given model can support cosmic acceleration, without solving for the on-shell solution. In this work, we present bounds on late-time FLRW-cosmologies in classes of theories that involve arbitrary numbers of sca…
▽ More
The cosmological dynamics of multiple scalar/pseudoscalar fields are difficult to solve, especially when the field-space metric is curved. This presents a challenge in determining whether a given model can support cosmic acceleration, without solving for the on-shell solution. In this work, we present bounds on late-time FLRW-cosmologies in classes of theories that involve arbitrary numbers of scalar and pseudoscalar fields coupled both kinetically (leading to a curved field space metric) and through scalar potentials. Such bounds are proven analytically, independently of initial conditions, with no approximation in the field equations and without referring to explicit solutions. Besides their broad applications to cosmological model building, our bounds can be applied to studying asymptotic cosmologies of certain classes of string compactifications.
△ Less
Submitted 24 June, 2024;
originally announced June 2024.
-
Regular algebras of dimension four associated to coordinate rings of rank-two quadrics
Authors:
R. G. Chandler,
H. Tran,
P. Veerapen,
X. Wang
Abstract:
In this paper, we classify connected graded quadratic Artin-Schelter regular (AS-regular, henceforth) algebras of global dimension four that have a Hilbert series the same as that of the polynomial ring on four generators and that map onto a twisted homogeneous coordinate ring of a rank-two quadric. A twisted homogeneous coordinate ring is a construction that was defined by Artin, Tate, and Van de…
▽ More
In this paper, we classify connected graded quadratic Artin-Schelter regular (AS-regular, henceforth) algebras of global dimension four that have a Hilbert series the same as that of the polynomial ring on four generators and that map onto a twisted homogeneous coordinate ring of a rank-two quadric. A twisted homogeneous coordinate ring is a construction that was defined by Artin, Tate, and Van den Bergh in \cite{ATV1, ATV2, AVdB1990} in the context of the classification of AS-regular algebras of global dimension three. In \cite{SV99,VVr}, the authors classified AS-regular algebras of global dimension four that map onto the twisted homogeneous coordinate ring of a rank-three and a rank-four quadric, respectively. We expand on their work to include the case of coordinate rings of a rank-two quadric.
△ Less
Submitted 18 June, 2024;
originally announced June 2024.
-
Convergence rates of S.O.S hierarchies for polynomial semidefinite programs
Authors:
Hoang Anh Tran,
Kim-Chuan Toh
Abstract:
We introduce a S.O.S hierarchy of lower bounds for a polynomial optimization problem whose constraint is expressed as a matrix polynomial semidefinite condition. Our approach involves utilizing a penalty function framework to directly address the matrix-based constraint, making it applicable to both discrete and continuous polynomial optimization problems. We investigate the convergence rates of t…
▽ More
We introduce a S.O.S hierarchy of lower bounds for a polynomial optimization problem whose constraint is expressed as a matrix polynomial semidefinite condition. Our approach involves utilizing a penalty function framework to directly address the matrix-based constraint, making it applicable to both discrete and continuous polynomial optimization problems. We investigate the convergence rates of these bounds in both problem types. The proposed method yields a variation of Putinar's theorem tailored for positive polynomials within a compact semidefinite set, defined by a matrix polynomial semidefinite constraint. More specifically, we derive novel insights into the convergence rates and degree of additional terms in the representation within this modified version of Putinar's theorem, based on the Jackson's theorem and a version of Łojasiewicz inequality.
△ Less
Submitted 17 June, 2024;
originally announced June 2024.
-
Reduced Basis method for finite volume simulations of parabolic PDEs applied to porous media flows
Authors:
Jana Tarhini,
Sébastien Boyaval,
Guillaume Enchéry,
Quang Huy Tran
Abstract:
Numerical simulations are a highly valuable tool to evaluate the impact of the uncertainties of various modelparameters, and to optimize e.g. injection-production scenarios in the context of underground storage (of CO2typically). Finite volume approximations of Darcy's parabolic model for flows in porous media are typically runmany times, for many values of parameters like permeability and porosit…
▽ More
Numerical simulations are a highly valuable tool to evaluate the impact of the uncertainties of various modelparameters, and to optimize e.g. injection-production scenarios in the context of underground storage (of CO2typically). Finite volume approximations of Darcy's parabolic model for flows in porous media are typically runmany times, for many values of parameters like permeability and porosity, at costly computational efforts.We study the relevance of reduced basis methods as a way to lower the overall simulation cost of finite volumeapproximations to Darcy's parabolic model for flows in porous media for different values of the parameters suchas permeability. In the context of underground gas storage (of CO2 typically) in saline aquifers, our aim isto evaluate quickly, for many parameter values, the flux along some interior boundaries near the well injectionarea-regarded as a quantity of interest-. To this end, we construct reduced bases by a standard POD-Greedyalgorithm. Our POD-Greedy algorithm uses a new goal-oriented error estimator designed from a discrete space-time energy norm independent of the parameter. We provide some numerical experiments that validate theefficiency of the proposed estimator.
△ Less
Submitted 12 June, 2024;
originally announced June 2024.
-
On isometry groups of gradient Ricci solitons
Authors:
Ha Tuan Dung,
Hung Tran
Abstract:
We give a result estimating the dimension of the Lie algebra of Killing vector fields on an irreducible non-trivial gradient Ricci soliton. Then we study the structure of this manifold when the maximal dimension is attained. There are local and global implications.
We give a result estimating the dimension of the Lie algebra of Killing vector fields on an irreducible non-trivial gradient Ricci soliton. Then we study the structure of this manifold when the maximal dimension is attained. There are local and global implications.
△ Less
Submitted 11 June, 2024;
originally announced June 2024.
-
Policy Iteration for Exploratory Hamilton--Jacobi--Bellman Equations
Authors:
Hung Vinh Tran,
Zhenhua Wang,
Yuming Paul Zhang
Abstract:
We study the policy iteration algorithm (PIA) for entropy-regularized stochastic control problems on an infinite time horizon with a large discount rate, focusing on two main scenarios. First, we analyze PIA with bounded coefficients where the controls applied to the diffusion term satisfy a smallness condition. We demonstrate the convergence of PIA based on a uniform $\mathcal{C}^{2,α}$ estimate…
▽ More
We study the policy iteration algorithm (PIA) for entropy-regularized stochastic control problems on an infinite time horizon with a large discount rate, focusing on two main scenarios. First, we analyze PIA with bounded coefficients where the controls applied to the diffusion term satisfy a smallness condition. We demonstrate the convergence of PIA based on a uniform $\mathcal{C}^{2,α}$ estimate for the value sequence generated by PIA, and provide a quantitative convergence analysis for this scenario. Second, we investigate PIA with unbounded coefficients but no control over the diffusion term. In this scenario, we first provide the well-posedness of the exploratory Hamilton--Jacobi--Bellman equation with linear growth coefficients and polynomial growth reward function. By such a well-posedess result we achieve PIA's convergence by establishing a quantitative locally uniform $\mathcal{C}^{1,α}$ estimates for the generated value sequence.
△ Less
Submitted 2 July, 2024; v1 submitted 2 June, 2024;
originally announced June 2024.
-
Regularity of normal Rees algebras of edge ideals of graphs
Authors:
Cao Huy Linh,
Quang Hoa Tran,
Thanh Vu
Abstract:
We classify all graphs for which the Rees algebras of their edge ideals are normal and have regularity equal to their matching numbers.
We classify all graphs for which the Rees algebras of their edge ideals are normal and have regularity equal to their matching numbers.
△ Less
Submitted 19 May, 2024;
originally announced May 2024.
-
Mahonian Statistics and Vincular Patterns on Permutations over Multisets
Authors:
Lien T. P. Ta,
Huong T. T. Tran
Abstract:
Most Mahonian statistics can be expressed as a linear combination of vincular patterns. This is not only true with statistics on the permutation set, but it can also be applied for statistics on the permutation with repetition set. By following the method extending the vincular patterns combinations presented by Kitaev and Vajnovszki, we discover 8 vincular-patterns combinations of mad and madl ex…
▽ More
Most Mahonian statistics can be expressed as a linear combination of vincular patterns. This is not only true with statistics on the permutation set, but it can also be applied for statistics on the permutation with repetition set. By following the method extending the vincular patterns combinations presented by Kitaev and Vajnovszki, we discover 8 vincular-patterns combinations of mad and madl extensions that are possible to be Mahonian. Some of these have been proved to be Mahonian on repetitive permutations by Clarke, Steingrimsson and Zeng, while the rest are new statistics extensions. In this thesis, we determine combinations of vincular pattern extension of mad and madl in Clarke, Steingrimsson and Zeng s paper, which have been proved to be Mahonian on the repetitive permutations. This result will be used to support the proof of Mahonity of the new statistics extensions. We show that these new statistics extensions are also Mahonian by constructing an involution Φ on repetitive permutations, which preserves the descents statistics and transforms new statistics extensions to Mahonian mad and madl extensions of Clarke.
△ Less
Submitted 14 May, 2024;
originally announced May 2024.
-
Quantitative homogenization of state-constraint Hamilton--Jacobi equations on perforated domains and applications
Authors:
Yuxi Han,
Wenjia **g,
Hiroyoshi Mitake,
Hung V. Tran
Abstract:
We study the periodic homogenization problem of state-constraint Hamilton--Jacobi equations on perforated domains in the convex setting and obtain the optimal convergence rate. We then consider a dilute situation in which the holes' diameter is much smaller than the microscopic scale. Finally, a homogenization problem with domain defects where some holes are missing is analyzed.
We study the periodic homogenization problem of state-constraint Hamilton--Jacobi equations on perforated domains in the convex setting and obtain the optimal convergence rate. We then consider a dilute situation in which the holes' diameter is much smaller than the microscopic scale. Finally, a homogenization problem with domain defects where some holes are missing is analyzed.
△ Less
Submitted 2 May, 2024;
originally announced May 2024.
-
Kähler Soliton Surfaces Are Generically Toric
Authors:
Hung Tran
Abstract:
Let $(M, g, ω, f, λ)$ be a Kähler gradient Ricci soliton in real dimension four. The first theorem states that it is an integrable Hamiltonian system in a classical sense. Furthermore, it is either of cohomogeneity one or the integrals of motion are given by the potential function $f$ and the scalar curvature $\text{S}$. The second theorem states that it must be toric under a generic assumption. T…
▽ More
Let $(M, g, ω, f, λ)$ be a Kähler gradient Ricci soliton in real dimension four. The first theorem states that it is an integrable Hamiltonian system in a classical sense. Furthermore, it is either of cohomogeneity one or the integrals of motion are given by the potential function $f$ and the scalar curvature $\text{S}$. The second theorem states that it must be toric under a generic assumption. That is, one assumes that the system is non-degenerate and the potential function $f$ is proper. In case $λ=0$, one further assumes that $f$ is bounded below or above. Then there is an effective, completely integrable Hamiltonian toric $\mathbb{T}^2$- action on $(M, ω)$.
△ Less
Submitted 29 April, 2024;
originally announced April 2024.
-
The Problem Of Image Super-Resolution, Denoising And Some Image Restoration Methods In Deep Learning Models
Authors:
Ngoc-Giau Pham,
Thanh-Hai Tong Le,
Van-Hieu Duong,
Hong-Ngoc Tran,
Phuoc-Hung Vo
Abstract:
In this article, we address the challenges of image super-resolution and noise reduction, which are crucial for enhancing the quality of images derived from low-resolution or noisy data. We compared and assessed several approaches for upgrading low-resolution images to higher resolutions and for eliminating unwanted noise, all while maintaining the essential characteristics of the original images…
▽ More
In this article, we address the challenges of image super-resolution and noise reduction, which are crucial for enhancing the quality of images derived from low-resolution or noisy data. We compared and assessed several approaches for upgrading low-resolution images to higher resolutions and for eliminating unwanted noise, all while maintaining the essential characteristics of the original images and recovering images from poor quality or damaged data using deep learning models. Our analysis and the experimental outcomes on image quality metrics indicate that the EDCNN neural network model, enhanced with pretrained weights, significantly outperforms other methods with a Train PSNR of 31.215, a Valid PSNR of 29.493, and a Test PSNR of 31.6632.
△ Less
Submitted 13 June, 2024; v1 submitted 15 April, 2024;
originally announced April 2024.
-
Closed $p$-Elastic Curves in Spheres of $\mathbb{L}^3$
Authors:
Alvaro Pampano,
Miraj Samarakkody,
Hung Tran
Abstract:
For every $p\in\mathbb{R}$, we study $p$-elastic curves in the hyperbolic plane $\mathbb{H}^2$ and in the de Sitter $2$-space $\mathbb{H}_1^2$. We analyze the existence of closed $p$-elastic curves with nonconstant curvature showing that in the hyperbolic plane $\mathbb{H}^2$ these curves exist provided that $p>1$, while in the de Sitter $2$-space $\mathbb{H}_1^2$ the restriction $p<0$ must be sat…
▽ More
For every $p\in\mathbb{R}$, we study $p$-elastic curves in the hyperbolic plane $\mathbb{H}^2$ and in the de Sitter $2$-space $\mathbb{H}_1^2$. We analyze the existence of closed $p$-elastic curves with nonconstant curvature showing that in the hyperbolic plane $\mathbb{H}^2$ these curves exist provided that $p>1$, while in the de Sitter $2$-space $\mathbb{H}_1^2$ the restriction $p<0$ must be satisfied.
△ Less
Submitted 12 April, 2024;
originally announced April 2024.
-
Shuffling Momentum Gradient Algorithm for Convex Optimization
Authors:
Trang H. Tran,
Quoc Tran-Dinh,
Lam M. Nguyen
Abstract:
The Stochastic Gradient Descent method (SGD) and its stochastic variants have become methods of choice for solving finite-sum optimization problems arising from machine learning and data science thanks to their ability to handle large-scale applications and big datasets. In the last decades, researchers have made substantial effort to study the theoretical performance of SGD and its shuffling vari…
▽ More
The Stochastic Gradient Descent method (SGD) and its stochastic variants have become methods of choice for solving finite-sum optimization problems arising from machine learning and data science thanks to their ability to handle large-scale applications and big datasets. In the last decades, researchers have made substantial effort to study the theoretical performance of SGD and its shuffling variants. However, only limited work has investigated its shuffling momentum variants, including shuffling heavy-ball momentum schemes for non-convex problems and Nesterov's momentum for convex settings. In this work, we extend the analysis of the shuffling momentum gradient method developed in [Tran et al (2021)] to both finite-sum convex and strongly convex optimization problems. We provide the first analysis of shuffling momentum-based methods for the strongly convex setting, attaining a convergence rate of $O(1/nT^2)$, where $n$ is the number of samples and $T$ is the number of training epochs. Our analysis is a state-of-the-art, matching the best rates of existing shuffling stochastic gradient algorithms in the literature.
△ Less
Submitted 5 March, 2024;
originally announced March 2024.
-
Hierarchy of the echo state property in quantum reservoir computing
Authors:
Shumpei Kobayashi,
Quoc Hoan Tran,
Kohei Nakajima
Abstract:
The echo state property (ESP) represents a fundamental concept in the reservoir computing (RC) framework that ensures output-only training of reservoir networks by being agnostic to the initial states and far past inputs. However, the traditional definition of ESP does not describe possible non-stationary systems in which statistical properties evolve. To address this issue, we introduce two new c…
▽ More
The echo state property (ESP) represents a fundamental concept in the reservoir computing (RC) framework that ensures output-only training of reservoir networks by being agnostic to the initial states and far past inputs. However, the traditional definition of ESP does not describe possible non-stationary systems in which statistical properties evolve. To address this issue, we introduce two new categories of ESP: $\textit{non-stationary ESP}$, designed for potentially non-stationary systems, and $\textit{subspace/subset ESP}$, designed for systems whose subsystems have ESP. Following the definitions, we numerically demonstrate the correspondence between non-stationary ESP in the quantum reservoir computer (QRC) framework with typical Hamiltonian dynamics and input encoding methods using non-linear autoregressive moving-average (NARMA) tasks. We also confirm the correspondence by computing linear/non-linear memory capacities that quantify input-dependent components within reservoir states. Our study presents a new understanding of the practical design of QRC and other possibly non-stationary RC systems in which non-stationary systems and subsystems are exploited.
△ Less
Submitted 17 April, 2024; v1 submitted 5 March, 2024;
originally announced March 2024.
-
Proximal Algorithms for a class of abstract convex functions
Authors:
Ewa Bednarczuk,
Dirk Lorenz,
The Hung Tran
Abstract:
In this paper we analyze a class of nonconvex optimization problem from the viewpoint of abstract convexity. Using the respective generalizations of the subgradient we propose an abstract notion proximal operator and derive a number of algorithms, namely an abstract proximal point method, an abstract forward-backward method and an abstract projected subgradient method. Global convergence results f…
▽ More
In this paper we analyze a class of nonconvex optimization problem from the viewpoint of abstract convexity. Using the respective generalizations of the subgradient we propose an abstract notion proximal operator and derive a number of algorithms, namely an abstract proximal point method, an abstract forward-backward method and an abstract projected subgradient method. Global convergence results for all algorithms are discussed and numerical examples are given
△ Less
Submitted 28 February, 2024; v1 submitted 26 February, 2024;
originally announced February 2024.
-
Stochastic ISTA/FISTA Adaptive Step Search Algorithms for Convex Composite Optimization
Authors:
Lam M. Nguyen,
Katya Scheinberg,
Trang H. Tran
Abstract:
We develop and analyze stochastic variants of ISTA and a full backtracking FISTA algorithms [Beck and Teboulle, 2009, Scheinberg et al., 2014] for composite optimization without the assumption that stochastic gradient is an unbiased estimator. This work extends analysis of inexact fixed step ISTA/FISTA in [Schmidt et al., 2011] to the case of stochastic gradient estimates and adaptive step-size pa…
▽ More
We develop and analyze stochastic variants of ISTA and a full backtracking FISTA algorithms [Beck and Teboulle, 2009, Scheinberg et al., 2014] for composite optimization without the assumption that stochastic gradient is an unbiased estimator. This work extends analysis of inexact fixed step ISTA/FISTA in [Schmidt et al., 2011] to the case of stochastic gradient estimates and adaptive step-size parameter chosen by backtracking. It also extends the framework for analyzing stochastic line-search method in [Cartis and Scheinberg, 2018] to the proximal gradient framework as well as to the accelerated first order methods.
△ Less
Submitted 23 February, 2024;
originally announced February 2024.
-
Optimal rate of convergence in periodic homogenization of viscous Hamilton-Jacobi equations
Authors:
Jianliang Qian,
Timo Sprekeler,
Hung V. Tran,
Yifeng Yu
Abstract:
We study the optimal rate of convergence in periodic homogenization of the viscous Hamilton-Jacobi equation $u^\varepsilon_t + H(\frac{x}{\varepsilon},Du^\varepsilon) = \varepsilon Δu^\varepsilon$ in $\mathbb R^n\times (0,\infty)$ subject to a given initial datum. We prove that $\|u^\varepsilon-u\|_{L^\infty(\mathbb R^n \times [0,T])} \leq C(1+T) \sqrt{\varepsilon}$ for any given $T>0$, where $u$…
▽ More
We study the optimal rate of convergence in periodic homogenization of the viscous Hamilton-Jacobi equation $u^\varepsilon_t + H(\frac{x}{\varepsilon},Du^\varepsilon) = \varepsilon Δu^\varepsilon$ in $\mathbb R^n\times (0,\infty)$ subject to a given initial datum. We prove that $\|u^\varepsilon-u\|_{L^\infty(\mathbb R^n \times [0,T])} \leq C(1+T) \sqrt{\varepsilon}$ for any given $T>0$, where $u$ is the viscosity solution of the effective problem. Moreover, we show that the $O(\sqrt{\varepsilon})$ rate is optimal for a natural class of $H$ and a Lipschitz continuous initial datum, both theoretically and through numerical experiments. It remains an interesting question to investigate whether the convergence rate can be improved when $H$ is uniformly convex. Finally, we propose a numerical scheme for the approximation of the effective Hamiltonian based on a finite element approximation of approximate corrector problems.
△ Less
Submitted 26 February, 2024; v1 submitted 5 February, 2024;
originally announced February 2024.
-
Quenching for axisymmetric hypersurfaces under forced mean curvature flows
Authors:
Hiroyoshi Mitake,
Yusuke Oka,
Hung Vinh Tran
Abstract:
Here, we study the motion of axisymmetric hypersurfaces $\{Γ_t\}_{t\ge0}$ evolved by forced mean curvature flows in the periodic setting. We establish conditions that quenching occurs or does not occur in terms of the initial data and forcing term. We also study the locations where the quenching happens in some special cases.
Here, we study the motion of axisymmetric hypersurfaces $\{Γ_t\}_{t\ge0}$ evolved by forced mean curvature flows in the periodic setting. We establish conditions that quenching occurs or does not occur in terms of the initial data and forcing term. We also study the locations where the quenching happens in some special cases.
△ Less
Submitted 10 January, 2024;
originally announced January 2024.
-
Asymptotic growth rate of solutions to level-set forced mean curvature flows with evolving spirals
Authors:
Hiroyoshi Mitake,
Hung V. Tran
Abstract:
Here, we study a level-set forced mean curvature flow with evolving spirals and the homogeneous Neumann boundary condition, which appears in a crystal growth model. Under some appropriate conditions on the forcing term, we prove that the solution is globally Lipschitz. We then study the large time average of the solution and deduce the asymptotic growth rate of the crystal. Some large time behavio…
▽ More
Here, we study a level-set forced mean curvature flow with evolving spirals and the homogeneous Neumann boundary condition, which appears in a crystal growth model. Under some appropriate conditions on the forcing term, we prove that the solution is globally Lipschitz. We then study the large time average of the solution and deduce the asymptotic growth rate of the crystal. Some large time behavior results of the solution are obtained.
△ Less
Submitted 13 December, 2023;
originally announced December 2023.
-
Collapsing universe before time
Authors:
Gary Shiu,
Flavio Tonioni,
Hung V. Tran
Abstract:
In this note, we prove analytic bounds on the equation of state of a cosmological fluid composed of an arbitrary number of canonical scalars evolving in a negative multi-exponential potential. Because of the negative energy, the universe is contracting and eventually undergoes a big crunch. A contracting universe is a fundamental feature of models of ekpyrosis and cyclic universes, which are a pro…
▽ More
In this note, we prove analytic bounds on the equation of state of a cosmological fluid composed of an arbitrary number of canonical scalars evolving in a negative multi-exponential potential. Because of the negative energy, the universe is contracting and eventually undergoes a big crunch. A contracting universe is a fundamental feature of models of ekpyrosis and cyclic universes, which are a proposed alternative to cosmic inflation to solve the flatness and horizon problems. Our analytic bounds set quantitative constraints on general effective theories of ekpyrosis. When applied to specific top-down constructions, our bounds can be used to determine whether ekpyrosis could in principle be realized. We point out some possible sources of tension in realizing the ekpyrotic universe in controlled constructions of string theory.
△ Less
Submitted 30 May, 2024; v1 submitted 11 December, 2023;
originally announced December 2023.
-
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.
-
Orthogonally weighted $\ell_{2,1}$ regularization for rank-aware joint sparse recovery: algorithm and analysis
Authors:
Armenak Petrosyan,
Konstantin Pieper,
Hoang Tran
Abstract:
We propose and analyze an efficient algorithm for solving the joint sparse recovery problem using a new regularization-based method, named orthogonally weighted $\ell_{2,1}$ ($\mathit{ow}\ell_{2,1}$), which is specifically designed to take into account the rank of the solution matrix. This method has applications in feature extraction, matrix column selection, and dictionary learning, and it is di…
▽ More
We propose and analyze an efficient algorithm for solving the joint sparse recovery problem using a new regularization-based method, named orthogonally weighted $\ell_{2,1}$ ($\mathit{ow}\ell_{2,1}$), which is specifically designed to take into account the rank of the solution matrix. This method has applications in feature extraction, matrix column selection, and dictionary learning, and it is distinct from commonly used $\ell_{2,1}$ regularization and other existing regularization-based approaches because it can exploit the full rank of the row-sparse solution matrix, a key feature in many applications. We provide a proof of the method's rank-awareness, establish the existence of solutions to the proposed optimization problem, and develop an efficient algorithm for solving it, whose convergence is analyzed. We also present numerical experiments to illustrate the theory and demonstrate the effectiveness of our method on real-life problems.
△ Less
Submitted 20 November, 2023;
originally announced November 2023.
-
Cutoff for the Glauber-Exclusion process in the full high-temperature regime: an information percolation approach
Authors:
Hong-Quan Tran
Abstract:
The Glauber-Exclusion process is a superposition of a Glauber dynamics and the Symmetric Simple Exclusion Process (SSEP) on the lattice. The model was shown to admit a reaction-diffusion equation as the hydrodynamic limit. In this article, we define a notion of temperature regimes via the reaction function in the equation and prove cutoff in the full high-temperature regime for the attractive mode…
▽ More
The Glauber-Exclusion process is a superposition of a Glauber dynamics and the Symmetric Simple Exclusion Process (SSEP) on the lattice. The model was shown to admit a reaction-diffusion equation as the hydrodynamic limit. In this article, we define a notion of temperature regimes via the reaction function in the equation and prove cutoff in the full high-temperature regime for the attractive model in dimensions $1$ and $2$ with periodic boundary condition. Our results show that the equation in the hydrodynamic limit reflects the mixing behavior of the large but finite system. Besides, cutoff is proved under the lack of reversibility and an explicit formula for the invariant measure. We also provide the spectral gap and prove pre-cutoff in all dimensions. Our proof involves a new interpretation of attractiveness, the information percolation framework introduced by Lubetzky and Sly, anti-concentration of simple random walk on the lattice, and a coupling inspired by excursion theory. We hope that this approach can find new applications in the future.
△ Less
Submitted 24 October, 2023;
originally announced October 2023.
-
The weak Lefschetz property of artinian algebras associated to paths and cycles
Authors:
Hop D. Nguyen,
Quang Hoa Tran
Abstract:
Given a base field $\Bbbk$ of characteristic zero, for each graph $G$, we associate the artinian algebra $A(G)$ defined by the edge ideal of $G$ and the squares of the variables. We study the weak Lefschetz property of $A(G)$. We classify some classes of graphs with relatively few edges, including paths and cycles, such that its associated artinian ring has the weak Lefschetz property.
Given a base field $\Bbbk$ of characteristic zero, for each graph $G$, we associate the artinian algebra $A(G)$ defined by the edge ideal of $G$ and the squares of the variables. We study the weak Lefschetz property of $A(G)$. We classify some classes of graphs with relatively few edges, including paths and cycles, such that its associated artinian ring has the weak Lefschetz property.
△ Less
Submitted 5 May, 2024; v1 submitted 22 October, 2023;
originally announced October 2023.
-
Kähler Solitons, Contact Structures, and Isoparametric Functions
Authors:
Hung Tran
Abstract:
Let $(M, g, f, J, λ)$ be a complete Kähler gradient Ricci soliton. Our first theorem classifies such a structure in real dimension four when $f$ has a geodesic gradient, a notion weaker than isoparametric. The soliton must be either a product metric or of cohomogeneity one with deformed homogeneous Sasakian orbits. The second result is a partially reverse statement in any dimension. Suppose that e…
▽ More
Let $(M, g, f, J, λ)$ be a complete Kähler gradient Ricci soliton. Our first theorem classifies such a structure in real dimension four when $f$ has a geodesic gradient, a notion weaker than isoparametric. The soliton must be either a product metric or of cohomogeneity one with deformed homogeneous Sasakian orbits. The second result is a partially reverse statement in any dimension. Suppose that each regular level set of $f$ is a deformed contact metric structure then the soliton is totally determined by a regular Sasakian model, which is a Riemannian submersion, with circle fibers, over a Kähler-Einstein manifold. In particular, $f$ must be isoparametric. An important ingredient of the proof is a characterization of a deformed Sasakian structure which generalizes a classical result.
△ Less
Submitted 17 October, 2023;
originally announced October 2023.
-
An optimal control approach for the treatment of hepatitis C patients
Authors:
Anh-Tuan Nguyen,
Hien Tran
Abstract:
In this article, the feasibility of using optimal control theory will be studied to develop control theoretic methods for personalized treatment of HCV patients. The mathematical model for HCV progression includes compartments for healthy hepatocytes, infected hepatocytes, infectious virions and noninfectious virions. Methodologies have been used from optimal control theory to design and synthesiz…
▽ More
In this article, the feasibility of using optimal control theory will be studied to develop control theoretic methods for personalized treatment of HCV patients. The mathematical model for HCV progression includes compartments for healthy hepatocytes, infected hepatocytes, infectious virions and noninfectious virions. Methodologies have been used from optimal control theory to design and synthesize an open-loop control based treatment regimen for HCV dynamics.
△ Less
Submitted 6 September, 2023; v1 submitted 5 September, 2023;
originally announced September 2023.
-
Spatial Invasion of Cooperative Parasites
Authors:
Vianney Brouard,
Cornelia Pokalyuk,
Marco Seiler,
Hung Tran
Abstract:
In this paper we study invasion probabilities and invasion times of cooperative parasites spreading in spatially structured host populations. The spatial structure of the host population is given by a random geometric graph on $[0,1]^n$, $n\in \mathbb{N}$, with a Poisson($N$)-distributed number of vertices and in which vertices are connected over an edge when they have a distance of at most…
▽ More
In this paper we study invasion probabilities and invasion times of cooperative parasites spreading in spatially structured host populations. The spatial structure of the host population is given by a random geometric graph on $[0,1]^n$, $n\in \mathbb{N}$, with a Poisson($N$)-distributed number of vertices and in which vertices are connected over an edge when they have a distance of at most $r_N\in Θ\left(N^{\frac{β-1}{n}}\right)$ for some $0<β<1$ and $N\rightarrow \infty$. At a host infection many parasites are generated and parasites move along edges to neighbouring hosts. We assume that parasites have to cooperate to infect hosts, in the sense that at least two parasites need to attack a host simultaneously. We find lower and upper bounds on the invasion probability of the parasites in terms of survival probabilities of branching processes with cooperation. Furthermore, we characterize the asymptotic invasion time.
An important ingredient of the proofs is a comparison with infection dynamics of cooperative parasites in host populations structured according to a complete graph, i.e. in well-mixed host populations. For these infection processes we can show that invasion probabilities are asymptotically equal to survival probabilities of branching processes with cooperation.
Furthermore, we build in the proofs on techniques developed in [BP22], where an analogous invasion process has been studied for host populations structured according to a configuration model.
We substantiate our results with simulations.
△ Less
Submitted 14 August, 2023;
originally announced August 2023.
-
Learning Networks from Gaussian Graphical Models and Gaussian Free Fields
Authors:
Subhro Ghosh,
Soumendu Sundar Mukherjee,
Hoang-Son Tran,
Ujan Gangopadhyay
Abstract:
We investigate the problem of estimating the structure of a weighted network from repeated measurements of a Gaussian Graphical Model (GGM) on the network. In this vein, we consider GGMs whose covariance structures align with the geometry of the weighted network on which they are based. Such GGMs have been of longstanding interest in statistical physics, and are referred to as the Gaussian Free Fi…
▽ More
We investigate the problem of estimating the structure of a weighted network from repeated measurements of a Gaussian Graphical Model (GGM) on the network. In this vein, we consider GGMs whose covariance structures align with the geometry of the weighted network on which they are based. Such GGMs have been of longstanding interest in statistical physics, and are referred to as the Gaussian Free Field (GFF). In recent years, they have attracted considerable interest in the machine learning and theoretical computer science. In this work, we propose a novel estimator for the weighted network (equivalently, its Laplacian) from repeated measurements of a GFF on the network, based on the Fourier analytic properties of the Gaussian distribution. In this pursuit, our approach exploits complex-valued statistics constructed from observed data, that are of interest on their own right. We demonstrate the effectiveness of our estimator with concrete recovery guarantees and bounds on the required sample complexity. In particular, we show that the proposed statistic achieves the parametric rate of estimation for fixed network size. In the setting of networks growing with sample size, our results show that for Erdos-Renyi random graphs $G(d,p)$ above the connectivity threshold, we demonstrate that network recovery takes place with high probability as soon as the sample size $n$ satisfies $n \gg d^4 \log d \cdot p^{-2}$.
△ Less
Submitted 4 August, 2023;
originally announced August 2023.
-
Nondivergence form degenerate linear parabolic equations on the upper half space
Authors:
Hongjie Dong,
Tuoc Phan,
Hung Vinh Tran
Abstract:
We study a class of nondivergence form second-order degenerate linear parabolic equations in $(-\infty, T) \times {\mathbb R}^d_+$ with the homogeneous Dirichlet boundary condition on $(-\infty, T) \times \partial {\mathbb R}^d_+$, where ${\mathbb R}^d_+ = \{x =(x_1,x_2,\ldots, x_d) \in {\mathbb R}^d\,:\, x_d>0\}$ and $T\in {(-\infty, \infty]}$ is given. The coefficient matrices of the equations a…
▽ More
We study a class of nondivergence form second-order degenerate linear parabolic equations in $(-\infty, T) \times {\mathbb R}^d_+$ with the homogeneous Dirichlet boundary condition on $(-\infty, T) \times \partial {\mathbb R}^d_+$, where ${\mathbb R}^d_+ = \{x =(x_1,x_2,\ldots, x_d) \in {\mathbb R}^d\,:\, x_d>0\}$ and $T\in {(-\infty, \infty]}$ is given. The coefficient matrices of the equations are the product of $μ(x_d)$ and bounded positive definite matrices, where $μ(x_d)$ behaves like $x_d^α$ for some given $α\in (0,2)$, which are degenerate on the boundary $\{x_d=0\}$ of the domain. The divergence form equations in this setting were studied in [14]. Under a partially weighted VMO assumption on the coefficients, we obtain the wellposedness and regularity of solutions in weighted Sobolev spaces. Our research program is motivated by the regularity theory of solutions to degenerate viscous Hamilton-Jacobi equations.
△ Less
Submitted 20 August, 2023; v1 submitted 20 June, 2023;
originally announced June 2023.
-
Late-time attractors and cosmic acceleration
Authors:
Gary Shiu,
Flavio Tonioni,
Hung V. Tran
Abstract:
We prove the conditions under which scaling cosmologies are inevitable late-time attractors of multi-field multi-exponential potentials, independently of initial conditions. The advantage of such scaling cosmologies is that the time dependence of the fields and of the scale factor is known analytically, thus allowing late-time observables to be determined exactly. Expanding the earlier results of…
▽ More
We prove the conditions under which scaling cosmologies are inevitable late-time attractors of multi-field multi-exponential potentials, independently of initial conditions. The advantage of such scaling cosmologies is that the time dependence of the fields and of the scale factor is known analytically, thus allowing late-time observables to be determined exactly. Expanding the earlier results of ref. arXiv:hep-th/2303.03418, here we continue the program of analytically characterizing the late-time behavior of cosmological solutions. Our results are general in that they are derived without relying on any approximation nor are they based on any assumption on the sources of the potential, such as their higher-dimensional or string-theoretic origin. We point out a number of model-independent features that follow from our analytic results, including a convex-hull criterion for cosmic acceleration. When applied to string theory, our analytic knowledge of late-time cosmological solutions enables us to single out potentials that can describe an accelerating universe from those which cannot and to quantitatively test several conjectured Swampland criteria.
△ Less
Submitted 22 September, 2023; v1 submitted 12 June, 2023;
originally announced June 2023.
-
Kähler Gradient Ricci Solitons with Large Symmetry
Authors:
Hung Tran
Abstract:
Let $(M, g, J, f)$ be an irreducible non-trivial Kähler gradient Ricci soliton of real dimension $2n$. We show that its group of isometries is of dimension at most $n^2$ and the case of equality is characterized. As a consequence, our framework shows the uniqueness of $U(n)$-invariant Kähler gradient Ricci solitons constructed earlier. There are corollaries regarding the groups of automorphisms or…
▽ More
Let $(M, g, J, f)$ be an irreducible non-trivial Kähler gradient Ricci soliton of real dimension $2n$. We show that its group of isometries is of dimension at most $n^2$ and the case of equality is characterized. As a consequence, our framework shows the uniqueness of $U(n)$-invariant Kähler gradient Ricci solitons constructed earlier. There are corollaries regarding the groups of automorphisms or affine transformations and a general version for almost Hermitian GRS. The approach is based on a connection to the geometry of an almost contact metric structure.
△ Less
Submitted 16 October, 2023; v1 submitted 9 June, 2023;
originally announced June 2023.
-
Analysis and Comparison of Two-Level KFAC Methods for Training Deep Neural Networks
Authors:
Abdoulaye Koroko,
Ani Anciaux-Sedrakian,
Ibtihel Ben Gharbia,
Valérie Garès,
Mounir Haddou,
Quang Huy Tran
Abstract:
As a second-order method, the Natural Gradient Descent (NGD) has the ability to accelerate training of neural networks. However, due to the prohibitive computational and memory costs of computing and inverting the Fisher Information Matrix (FIM), efficient approximations are necessary to make NGD scalable to Deep Neural Networks (DNNs). Many such approximations have been attempted. The most sophis…
▽ More
As a second-order method, the Natural Gradient Descent (NGD) has the ability to accelerate training of neural networks. However, due to the prohibitive computational and memory costs of computing and inverting the Fisher Information Matrix (FIM), efficient approximations are necessary to make NGD scalable to Deep Neural Networks (DNNs). Many such approximations have been attempted. The most sophisticated of these is KFAC, which approximates the FIM as a block-diagonal matrix, where each block corresponds to a layer of the neural network. By doing so, KFAC ignores the interactions between different layers. In this work, we investigate the interest of restoring some low-frequency interactions between the layers by means of two-level methods. Inspired from domain decomposition, several two-level corrections to KFAC using different coarse spaces are proposed and assessed. The obtained results show that incorporating the layer interactions in this fashion does not really improve the performance of KFAC. This suggests that it is safe to discard the off-diagonal blocks of the FIM, since the block-diagonal approach is sufficiently robust, accurate and economical in computation time.
△ Less
Submitted 3 April, 2023; v1 submitted 31 March, 2023;
originally announced March 2023.
-
Well-Rounded ideal lattices of cyclic cubic and quartic fields
Authors:
Dat T. Tran,
Nam H. Le,
Ha T. N. Tran
Abstract:
In this paper, we find criteria for when cyclic cubic and cyclic quartic fields have well-rounded ideal lattices. We show that every cyclic cubic field has at least one well-rounded ideal. We also prove that there exist families of cyclic quartic fields which have well-rounded ideals and explicitly construct their minimal bases. In addition, for a given prime number $p$, if a cyclic quartic field…
▽ More
In this paper, we find criteria for when cyclic cubic and cyclic quartic fields have well-rounded ideal lattices. We show that every cyclic cubic field has at least one well-rounded ideal. We also prove that there exist families of cyclic quartic fields which have well-rounded ideals and explicitly construct their minimal bases. In addition, for a given prime number $p$, if a cyclic quartic field has a unique prime ideal above $p$, then we provide the necessary and sufficient conditions for that ideal to be well-rounded. Moreover, in cyclic quartic fields, we provide the prime decomposition of all odd prime numbers and construct an explicit integral basis for every prime ideal.
△ Less
Submitted 13 October, 2023; v1 submitted 29 March, 2023;
originally announced March 2023.
-
Bifurcation of homogenization and nonhomogenization of the curvature G-equation with shear flows
Authors:
Hiroyoshi Mitake,
Connor Mooney,
Hung V. Tran,
Jack Xin,
Yifeng Yu
Abstract:
The level-set curvature G-equation, a well-known model in turbulent combustion, has the following form
$G_t + \left(1-d\, \mathrm{dvi}\left({\frac{DG}{|DG|}}\right)\right)_+|DG|+V(X)\cdot DG=0.$
Here the cutoff correction $()_+$ is imposed to avoid non-physical negative local burning velocity. The existence of the effective burning velocity has been established for a large class of physically…
▽ More
The level-set curvature G-equation, a well-known model in turbulent combustion, has the following form
$G_t + \left(1-d\, \mathrm{dvi}\left({\frac{DG}{|DG|}}\right)\right)_+|DG|+V(X)\cdot DG=0.$
Here the cutoff correction $()_+$ is imposed to avoid non-physical negative local burning velocity. The existence of the effective burning velocity has been established for a large class of physically relevant incompressible flows $V$ in two dimensions [13] via game theory dynamics. In this paper, we show that the effective burning velocity associated with shear flows in dimensions three or higher ceases to exist when the flow intensity surpasses a bifurcation point. The characterization of the bifurcation point in three dimensions is closely related to the regularity theory of two-dimensional minimal surface type equations due to [29]. As a consequence, a bifurcation also exists for the validity of full homogenization of the curvature G-equation associated with shear flows.
△ Less
Submitted 28 March, 2023;
originally announced March 2023.
-
Forward-Backward algorithms for weakly convex problems
Authors:
Ewa Bednarczuk,
Giovanni Bruccola,
Gabriele Scrivanti,
The Hung Tran
Abstract:
We investigate the convergence properties of exact and inexact forward-backward algorithms to minimise the sum of two weakly convex functions defined on a Hilbert space, where one has a Lipschitz-continuous gradient. We show that the exact forward-backward algorithm converges strongly to a global solution, provided that the objective function satisfies a sharpness condition. For the inexact forwar…
▽ More
We investigate the convergence properties of exact and inexact forward-backward algorithms to minimise the sum of two weakly convex functions defined on a Hilbert space, where one has a Lipschitz-continuous gradient. We show that the exact forward-backward algorithm converges strongly to a global solution, provided that the objective function satisfies a sharpness condition. For the inexact forward-backward algorithm, the same condition ensures that the distance from the iterates to the solution set approaches a positive threshold depending on the accuracy level of the proximal computations. As an application of the considered setting, we provide numerical experiments related to discrete tomography.
△ Less
Submitted 21 June, 2024; v1 submitted 24 March, 2023;
originally announced March 2023.
-
Accelerating universe at the end of time
Authors:
Gary Shiu,
Flavio Tonioni,
Hung V. Tran
Abstract:
We investigate whether an accelerating universe can be realized as an asymptotic late-time solution of FLRW-cosmology with multi-field multi-exponential potentials. Late-time cosmological solutions exhibit a universal behavior which enables us to bound the rate of time variation of the Hubble parameter. In string-theoretic realizations, if the dilaton remains a rolling field, our bound singles out…
▽ More
We investigate whether an accelerating universe can be realized as an asymptotic late-time solution of FLRW-cosmology with multi-field multi-exponential potentials. Late-time cosmological solutions exhibit a universal behavior which enables us to bound the rate of time variation of the Hubble parameter. In string-theoretic realizations, if the dilaton remains a rolling field, our bound singles out a tension in achieving asymptotic late-time cosmic acceleration. Our findings go beyond previous no-go theorems in that they apply to arbitrary multi-exponential potentials and make no specific reference to vacuum or slow-roll solutions. We also show that if the late-time solution approaches a critical point of the dynamical system governing the cosmological evolution, the criterion for cosmic acceleration can be generally stated in terms of a directional derivative of the potential.
△ Less
Submitted 21 September, 2023; v1 submitted 6 March, 2023;
originally announced March 2023.
-
Convergence analysis for a nonlocal gradient descent method via directional Gaussian smoothing
Authors:
Hoang Tran,
Qiang Du,
Guannan Zhang
Abstract:
We analyze the convergence of a nonlocal gradient descent method for minimizing a class of high-dimensional non-convex functions, where a directional Gaussian smoothing (DGS) is proposed to define the nonlocal gradient (also referred to as the DGS gradient). The method was first proposed in [42], in which multiple numerical experiments showed that replacing the traditional local gradient with the…
▽ More
We analyze the convergence of a nonlocal gradient descent method for minimizing a class of high-dimensional non-convex functions, where a directional Gaussian smoothing (DGS) is proposed to define the nonlocal gradient (also referred to as the DGS gradient). The method was first proposed in [42], in which multiple numerical experiments showed that replacing the traditional local gradient with the DGS gradient can help the optimizers escape local minima more easily and significantly improve their performance. However, a rigorous theory for the efficiency of the method on nonconvex landscape is lacking. In this work, we investigate the scenario where the objective function is composed of a convex function, perturbed by a oscillating noise. We provide a convergence theory under which the iterates exponentially converge to a tightened neighborhood of the solution, whose size is characterized by the noise wavelength. We also establish a correlation between the optimal values of the Gaussian smoothing radius and the noise wavelength, thus justify the advantage of using moderate or large smoothing radius with the method. Furthermore, if the noise level decays to zero when approaching global minimum, we prove that DGS-based optimization converges to the exact global minimum with linear rates, similarly to standard gradient-based method in optimizing convex functions. Several numerical experiments are provided to confirm our theory and illustrate the superiority of the approach over those based on the local gradient.
△ Less
Submitted 13 February, 2023;
originally announced February 2023.
-
The size function for imaginary cyclic sextic fields
Authors:
Ha Thanh Nguyen Tran,
Peng Tian,
Amy Feaver
Abstract:
In this paper, we investigate the size function $h^0$ for number fields. This size function is analogous to the dimension of the Riemann-Roch spaces of divisors on an algebraic curve. Van der Geer and Schoof conjectured that $h^0$ attains its maximum at the trivial class of Arakelov divisors. This conjecture was proved for all number fields with unit group of rank 0 and 1, and also for cyclic cubi…
▽ More
In this paper, we investigate the size function $h^0$ for number fields. This size function is analogous to the dimension of the Riemann-Roch spaces of divisors on an algebraic curve. Van der Geer and Schoof conjectured that $h^0$ attains its maximum at the trivial class of Arakelov divisors. This conjecture was proved for all number fields with unit group of rank 0 and 1, and also for cyclic cubic fields which have unit group of rank two. We prove the conjecture also holds for totally imaginary cyclic sextic fields, another class of number fields with unit group of rank two.
△ Less
Submitted 12 February, 2023;
originally announced February 2023.
-
Optimal convergence rates in stochastic homogenization in a balanced random environment
Authors:
Xiaoqin Guo,
Hung V. Tran
Abstract:
We consider random walks in a uniformly elliptic, balanced, i.i.d. random environment in the integer lattice $Z^d$ for $d\geq 2$ and the corresponding problem of stochastic homogenization of non-divergence form difference operators. We first derive a quantitative law of large numbers for the invariant measure, which is nearly optimal. A mixing property of the field of the invariant measure is then…
▽ More
We consider random walks in a uniformly elliptic, balanced, i.i.d. random environment in the integer lattice $Z^d$ for $d\geq 2$ and the corresponding problem of stochastic homogenization of non-divergence form difference operators. We first derive a quantitative law of large numbers for the invariant measure, which is nearly optimal. A mixing property of the field of the invariant measure is then achieved. We next obtain rates of convergence for the homogenization of the Dirichlet problem for non-divergence form operators, which are generically optimal for $d\geq 3$ and nearly optimal when $d=2$. Furthermore, we establish the existence, stationarity and uniqueness properties of the corrector problem for all dimensions $d\ge 2$. Afterwards, we quantify the ergodicity of the environmental process for both the continuous-time and discrete-time random walks, and as a consequence, we get explicit convergence rates for the quenched central limit theorem of the balanced random walk.
△ Less
Submitted 13 April, 2023; v1 submitted 3 January, 2023;
originally announced January 2023.
-
Policy iteration for the deterministic control problems -- a viscosity approach
Authors:
Wenpin Tang,
Hung Vinh Tran,
Yuming Paul Zhang
Abstract:
This paper is concerned with the convergence rate of policy iteration for (deterministic) optimal control problems in continuous time. To overcome the problem of ill-posedness due to lack of regularity, we consider a semi-discrete scheme by adding a viscosity term via finite differences in space. We prove that PI for the semi-discrete scheme converges exponentially fast, and provide a bound on the…
▽ More
This paper is concerned with the convergence rate of policy iteration for (deterministic) optimal control problems in continuous time. To overcome the problem of ill-posedness due to lack of regularity, we consider a semi-discrete scheme by adding a viscosity term via finite differences in space. We prove that PI for the semi-discrete scheme converges exponentially fast, and provide a bound on the error induced by the semi-discrete scheme. We also consider the discrete space-time scheme, where both space and time are discretized. Convergence rate of PI and the discretization error are studied.
△ Less
Submitted 1 January, 2023;
originally announced January 2023.
-
Asymptotic regularity of invariant chains of edge ideals
Authors:
Do Trong Hoang,
Hop D. Nguyen,
Quang Hoa Tran
Abstract:
We study chains of nonzero edge ideals that are invariant under the action of the monoid $\mathrm{Inc}$ of increasing functions on the positive integers. We prove that the sequence of Castelnuovo--Mumford regularity of ideals in such a chain is eventually constant with limit either 2 or 3, and we determine explicitly when the constancy behaviour sets in. This provides further evidence to a conject…
▽ More
We study chains of nonzero edge ideals that are invariant under the action of the monoid $\mathrm{Inc}$ of increasing functions on the positive integers. We prove that the sequence of Castelnuovo--Mumford regularity of ideals in such a chain is eventually constant with limit either 2 or 3, and we determine explicitly when the constancy behaviour sets in. This provides further evidence to a conjecture on the asymptotic linearity of the regularity of $\mathrm{Inc}$-invariant chains of homogeneous ideals. The proofs reveal unexpected combinatorial properties of $\mathrm{Inc}$-invariant chains of edge ideals.
△ Less
Submitted 28 January, 2023; v1 submitted 13 December, 2022;
originally announced December 2022.
-
Rigidity of four-dimensional Kähler-Ricci solitons
Authors:
Xiaodong Cao,
Ernani Ribeiro Jr,
Hung Tran
Abstract:
In this article, we investigate four-dimensional gradient shrinking Ricci solitons close to a Kähler model. The first theorem could be considered as a rigidity result for the Kähler-Ricci soliton structure on $\mathbb{S}^2\times \mathbb{R}^2$ (in the sense of Remark 1). Moreover, we show that if the quotient of norm of the self-dual Weyl tensor and scalar curvature is close to that on a Kähler met…
▽ More
In this article, we investigate four-dimensional gradient shrinking Ricci solitons close to a Kähler model. The first theorem could be considered as a rigidity result for the Kähler-Ricci soliton structure on $\mathbb{S}^2\times \mathbb{R}^2$ (in the sense of Remark 1). Moreover, we show that if the quotient of norm of the self-dual Weyl tensor and scalar curvature is close to that on a Kähler metric in a specific sense, then the gradient Ricci soliton must be either half-conformally flat or locally Kähler.
△ Less
Submitted 10 December, 2022;
originally announced December 2022.
-
Cutoff for the non reversible SSEP with reservoirs
Authors:
Hong-Quan Tran
Abstract:
We consider the Symmetric Simple Exclusion Process (SSEP) on the segment with two reservoirs of densities $p, q \in (0,1)$ at the two endpoints. We show that the system exhibits cutoff with a diffusive window, thus confirming a conjecture of Gantert, Nestoridi, and Schmid in \cite{Gantert2020}. In particular, our result covers the regime $p \neq q$, where the process is not reversible and there is…
▽ More
We consider the Symmetric Simple Exclusion Process (SSEP) on the segment with two reservoirs of densities $p, q \in (0,1)$ at the two endpoints. We show that the system exhibits cutoff with a diffusive window, thus confirming a conjecture of Gantert, Nestoridi, and Schmid in \cite{Gantert2020}. In particular, our result covers the regime $p \neq q$, where the process is not reversible and there is no known explicit formula for the invariant measure. Our proof exploits the information percolation framework introduced by Lubetzky and Sly, the negative dependence of the system, and an anticoncentration inequality at the conditional level. We believe this approach is applicable to other models.
△ Less
Submitted 26 November, 2022;
originally announced November 2022.
-
Calculus rules for proximal ε-subdifferentials and inexact proximity operators for weakly convex functions
Authors:
Ewa Bednarczuk,
Giovanni Bruccola,
Gabriele Scrivanti,
The Hung Tran
Abstract:
We investigate inexact proximity operators for weakly convex functions. To this aim, we derive sum rules for proximal ε-subdifferentials, by incorporating the moduli of weak convexity of the functions into the respective formulas. This allows us to investigate inexact proximity operators for weakly convex functions in terms of proximal ε-subdifferentials.
We investigate inexact proximity operators for weakly convex functions. To this aim, we derive sum rules for proximal ε-subdifferentials, by incorporating the moduli of weak convexity of the functions into the respective formulas. This allows us to investigate inexact proximity operators for weakly convex functions in terms of proximal ε-subdifferentials.
△ Less
Submitted 23 April, 2024; v1 submitted 26 November, 2022;
originally announced November 2022.
-
A new proof of Stanley's theorem on the strong Lefschetz property
Authors:
Ho V. N. Phuong,
Quang Hoa Tran
Abstract:
A standard graded artinian monomial complete intersection algebra $A=\Bbbk[x_1,x_2,\ldots,x_n]/(x_1^{a_1},x_2^{a_2},\ldots,x_n^{a_n})$, with $\Bbbk$ a field of characteristic zero, has the strong Lefschetz property due to Stanley in 1980. In this paper, we give a new proof for this result by using only the basic properties of linear algebra. Furthermore, our proof is still true in the case where t…
▽ More
A standard graded artinian monomial complete intersection algebra $A=\Bbbk[x_1,x_2,\ldots,x_n]/(x_1^{a_1},x_2^{a_2},\ldots,x_n^{a_n})$, with $\Bbbk$ a field of characteristic zero, has the strong Lefschetz property due to Stanley in 1980. In this paper, we give a new proof for this result by using only the basic properties of linear algebra. Furthermore, our proof is still true in the case where the characteristic of $\Bbbk$ is greater than the socle degree of $A$, namely $a_1+a_2+\cdots+a_n - n$.
△ Less
Submitted 24 November, 2022;
originally announced November 2022.
-
The Generalized Elastic Net for least squares regression with network-aligned signal and correlated design
Authors:
Huy Tran,
Sansen Wei,
Claire Donnat
Abstract:
We propose a novel $\ell_1+\ell_2$-penalty, which we refer to as the Generalized Elastic Net, for regression problems where the feature vectors are indexed by vertices of a given graph and the true signal is believed to be smooth or piecewise constant with respect to this graph. Under the assumption of correlated Gaussian design, we derive upper bounds for the prediction and estimation errors, whi…
▽ More
We propose a novel $\ell_1+\ell_2$-penalty, which we refer to as the Generalized Elastic Net, for regression problems where the feature vectors are indexed by vertices of a given graph and the true signal is believed to be smooth or piecewise constant with respect to this graph. Under the assumption of correlated Gaussian design, we derive upper bounds for the prediction and estimation errors, which are graph-dependent and consist of a parametric rate for the unpenalized portion of the regression vector and another term that depends on our network alignment assumption. We also provide a coordinate descent procedure based on the Lagrange dual objective to compute this estimator for large-scale problems. Finally, we compare our proposed estimator to existing regularized estimators on a number of real and synthetic datasets and discuss its potential limitations.
△ Less
Submitted 24 November, 2022; v1 submitted 1 November, 2022;
originally announced November 2022.
-
Well-Rounded Twists of Ideal Lattices from Imaginary Quadratic Fields
Authors:
Nam H. Le,
Dat T. Tran,
Ha T. N. Tran
Abstract:
In this paper, we investigate the properties of well-rounded twists of a given ideal lattice of an imaginary quadratic field $K$. We show that every ideal lattice $I$ of $K$ has at least one well-rounded twist lattice. Moreover, we provide an explicit algorithm to compute all well-rounded twists of $I$.
In this paper, we investigate the properties of well-rounded twists of a given ideal lattice of an imaginary quadratic field $K$. We show that every ideal lattice $I$ of $K$ has at least one well-rounded twist lattice. Moreover, we provide an explicit algorithm to compute all well-rounded twists of $I$.
△ Less
Submitted 26 October, 2022;
originally announced October 2022.
-
Stochastic integrability of heat-kernel bounds for random walks in a balanced random environment
Authors:
Xiaoqin Guo,
Hung V. Tran
Abstract:
We consider random walks in a balanced i.i.d. random environment in $Z^d$ for $d\ge2$ and the corresponding discrete non-divergence form difference operators. We first obtain an exponential integrability of the heat kernel bounds. We then prove the optimal diffusive decay of the semigroup generated by the heat kernel for $d\ge3$. As a consequence, we deduce a functional central limit theorem for t…
▽ More
We consider random walks in a balanced i.i.d. random environment in $Z^d$ for $d\ge2$ and the corresponding discrete non-divergence form difference operators. We first obtain an exponential integrability of the heat kernel bounds. We then prove the optimal diffusive decay of the semigroup generated by the heat kernel for $d\ge3$. As a consequence, we deduce a functional central limit theorem for the environment viewed from the particle.
△ Less
Submitted 28 September, 2022;
originally announced September 2022.
