-
Improved convergence rates for the Hele-Shaw limit in the presence of confining potentials
Authors:
Noemi David,
Alpár R. Mészáros,
Filippo Santambrogio
Abstract:
Nowadays a vast literature is available on the Hele-Shaw or incompressible limit for nonlinear degenerate diffusion equations. This problem has attracted a lot of attention due to its applications to tissue growth and crowd motion modelling as it constitutes a way to link soft congestion (or compressible) models to hard congestion (or incompressible) descriptions. In this paper, we address the que…
▽ More
Nowadays a vast literature is available on the Hele-Shaw or incompressible limit for nonlinear degenerate diffusion equations. This problem has attracted a lot of attention due to its applications to tissue growth and crowd motion modelling as it constitutes a way to link soft congestion (or compressible) models to hard congestion (or incompressible) descriptions. In this paper, we address the question of estimating the rate of this asymptotics in the presence of external drifts. In particular, we provide improved results in the 2-Wasserstein distance which are global in time thanks to the contractivity property that holds for strictly convex potentials.
△ Less
Submitted 12 May, 2024;
originally announced May 2024.
-
Position: Understanding LLMs Requires More Than Statistical Generalization
Authors:
Patrik Reizinger,
Szilvia Ujváry,
Anna Mészáros,
Anna Kerekes,
Wieland Brendel,
Ferenc Huszár
Abstract:
The last decade has seen blossoming research in deep learning theory attempting to answer, "Why does deep learning generalize?" A powerful shift in perspective precipitated this progress: the study of overparametrized models in the interpolation regime. In this paper, we argue that another perspective shift is due, since some of the desirable qualities of LLMs are not a consequence of good statist…
▽ More
The last decade has seen blossoming research in deep learning theory attempting to answer, "Why does deep learning generalize?" A powerful shift in perspective precipitated this progress: the study of overparametrized models in the interpolation regime. In this paper, we argue that another perspective shift is due, since some of the desirable qualities of LLMs are not a consequence of good statistical generalization and require a separate theoretical explanation. Our core argument relies on the observation that AR probabilistic models are inherently non-identifiable: models zero or near-zero KL divergence apart -- thus, equivalent test loss -- can exhibit markedly different behaviors. We support our position with mathematical examples and empirical observations, illustrating why non-identifiability has practical relevance through three case studies: (1) the non-identifiability of zero-shot rule extrapolation; (2) the approximate non-identifiability of in-context learning; and (3) the non-identifiability of fine-tunability. We review promising research directions focusing on LLM-relevant generalization measures, transferability, and inductive biases.
△ Less
Submitted 17 June, 2024; v1 submitted 3 May, 2024;
originally announced May 2024.
-
Convergence analysis of controlled particle systems arising in deep learning: from finite to infinite sample size
Authors:
Huafu Liao,
Alpár R. Mészáros,
Chenchen Mou,
Chao Zhou
Abstract:
This paper deals with a class of neural SDEs and studies the limiting behavior of the associated sampled optimal control problems as the sample size grows to infinity. The neural SDEs with N samples can be linked to the N-particle systems with centralized control. We analyze the Hamilton--Jacobi--Bellman equation corresponding to the N-particle system and establish regularity results which are uni…
▽ More
This paper deals with a class of neural SDEs and studies the limiting behavior of the associated sampled optimal control problems as the sample size grows to infinity. The neural SDEs with N samples can be linked to the N-particle systems with centralized control. We analyze the Hamilton--Jacobi--Bellman equation corresponding to the N-particle system and establish regularity results which are uniform in N. The uniform regularity estimates are obtained by the stochastic maximum principle and the analysis of a backward stochastic Riccati equation. Using these uniform regularity results, we show the convergence of the minima of objective functionals and optimal parameters of the neural SDEs as the sample size N tends to infinity. The limiting objects can be identified with suitable functions defined on the Wasserstein space of Borel probability measures. Furthermore, quantitative algebraic convergence rates are also obtained.
△ Less
Submitted 8 April, 2024;
originally announced April 2024.
-
The $2$-torsion of determinantal hypertrees is not Cohen-Lenstra
Authors:
András Mészáros
Abstract:
Let $T_n$ be a $2$-dimensional determinantal hypertree on $n$ vertices. Kahle and Newman conjectured that the $p$-torsion of $H_1(T_n,\mathbb{Z})$ asymptotically follows the Cohen-Lenstra distribution. For $p=2$, we disprove this conjecture by showing that given a positive integer $h$, for all large enough $n$, we have \[\mathbb{P}(\dim H_1(T_n,\mathbb{F}_2)\ge h)\ge \frac{e^{-200h}}{(100h)^{5h}}.…
▽ More
Let $T_n$ be a $2$-dimensional determinantal hypertree on $n$ vertices. Kahle and Newman conjectured that the $p$-torsion of $H_1(T_n,\mathbb{Z})$ asymptotically follows the Cohen-Lenstra distribution. For $p=2$, we disprove this conjecture by showing that given a positive integer $h$, for all large enough $n$, we have \[\mathbb{P}(\dim H_1(T_n,\mathbb{F}_2)\ge h)\ge \frac{e^{-200h}}{(100h)^{5h}}.\] We also show that $T_n$ is a bad cosystolic expander with positive probability.
△ Less
Submitted 2 April, 2024;
originally announced April 2024.
-
Kinetic-type Mean Field Games with Non-separable Local Hamiltonians
Authors:
David M. Ambrose,
Megan Griffin-Pickering,
Alpár R. Mészáros
Abstract:
We prove well-posedness of a class of kinetic-type Mean Field Games, which typically arise when agents control their acceleration. Such systems include independent variables representing the spatial position as well as velocity. We consider non-separable Hamiltonians without any structural conditions, which depend locally on the density variable. Our analysis is based on two main ingredients: an e…
▽ More
We prove well-posedness of a class of kinetic-type Mean Field Games, which typically arise when agents control their acceleration. Such systems include independent variables representing the spatial position as well as velocity. We consider non-separable Hamiltonians without any structural conditions, which depend locally on the density variable. Our analysis is based on two main ingredients: an energy method for the forward-backward system in Sobolev spaces, on the one hand and on a suitable vector field method to control derivatives with respect to the velocity variable, on the other hand. The careful combination of these two techniques reveals interesting phenomena applicable for Mean Field Games involving general classes of drift-diffusion operators and nonlinearities. While many prior existence theories for general Mean Field Games systems take the final datum function to be smoothing, we can allow this function to be non-smoothing, i.e. also depending locally on the final measure. Our well-posedness results hold under an appropriate smallness condition, assumed jointly on the data.
△ Less
Submitted 19 March, 2024;
originally announced March 2024.
-
Hidden monotonicity and canonical transformations for mean field games and master equations
Authors:
Mohit Bansil,
Alpár R. Mészáros
Abstract:
In this paper we unveil novel monotonicity conditions applicable for Mean Field Games through the exploration of finite dimensional $canonical\ transformations$. Our findings contribute to establishing new global well-posedness results for the associated master equations, also in the case of potentially degenerate idiosyncratic noise. Additionally, we show that recent advancements in global well-p…
▽ More
In this paper we unveil novel monotonicity conditions applicable for Mean Field Games through the exploration of finite dimensional $canonical\ transformations$. Our findings contribute to establishing new global well-posedness results for the associated master equations, also in the case of potentially degenerate idiosyncratic noise. Additionally, we show that recent advancements in global well-posedness results, specifically those related to displacement semi-monotone and anti-monotone data, can be easily obtained as a consequence of our main results.
△ Less
Submitted 8 March, 2024;
originally announced March 2024.
-
On classical solutions and canonical transformations for Hamilton--Jacobi--Bellman equations
Authors:
Mohit Bansil,
Alpár R. Mészáros
Abstract:
In this note we show how canonical transformations reveal hidden convexity properties for deterministic optimal control problems, which in turn result in global existence of $C^{1,1}_{loc}$ solutions to first order Hamilton--Jacobi--Bellman equations.
In this note we show how canonical transformations reveal hidden convexity properties for deterministic optimal control problems, which in turn result in global existence of $C^{1,1}_{loc}$ solutions to first order Hamilton--Jacobi--Bellman equations.
△ Less
Submitted 8 March, 2024;
originally announced March 2024.
-
Bounds on the mod 2 homology of random 2-dimensional simplicial complexes
Authors:
András Mészáros
Abstract:
As a first step towards a conjecture of Kahle and Newman, we prove that if $T_n$ is a random $2$-dimensional determinantal hypertree on $n$ vertices, then \[\frac{\dim H_1(T_n,\mathbb{F}_2)}{n^2}\] converges to zero in probability.
Confirming a conjecture of Linial and Peled, we also prove the analogous statement for the $1$-out $2$-complex.
Our proof relies on the large deviation principle fo…
▽ More
As a first step towards a conjecture of Kahle and Newman, we prove that if $T_n$ is a random $2$-dimensional determinantal hypertree on $n$ vertices, then \[\frac{\dim H_1(T_n,\mathbb{F}_2)}{n^2}\] converges to zero in probability.
Confirming a conjecture of Linial and Peled, we also prove the analogous statement for the $1$-out $2$-complex.
Our proof relies on the large deviation principle for the Erdős-Rényi random graph by Chatterjee and Varadhan.
△ Less
Submitted 24 January, 2024;
originally announced January 2024.
-
Robust Multi-Modal Density Estimation
Authors:
Anna Mészáros,
Julian F. Schumann,
Javier Alonso-Mora,
Arkady Zgonnikov,
Jens Kober
Abstract:
The estimation of probability density functions is a fundamental problem in science and engineering. However, common methods such as kernel density estimation (KDE) have been demonstrated to lack robustness, while more complex methods have not been evaluated in multi-modal estimation problems. In this paper, we present ROME (RObust Multi-modal Estimator), a non-parametric approach for density esti…
▽ More
The estimation of probability density functions is a fundamental problem in science and engineering. However, common methods such as kernel density estimation (KDE) have been demonstrated to lack robustness, while more complex methods have not been evaluated in multi-modal estimation problems. In this paper, we present ROME (RObust Multi-modal Estimator), a non-parametric approach for density estimation which addresses the challenge of estimating multi-modal, non-normal, and highly correlated distributions. ROME utilizes clustering to segment a multi-modal set of samples into multiple uni-modal ones and then combines simple KDE estimates obtained for individual clusters in a single multi-modal estimate. We compared our approach to state-of-the-art methods for density estimation as well as ablations of ROME, showing that it not only outperforms established methods but is also more robust to a variety of distributions. Our results demonstrate that ROME can overcome the issues of over-fitting and over-smoothing exhibited by other estimators.
△ Less
Submitted 6 May, 2024; v1 submitted 19 January, 2024;
originally announced January 2024.
-
Coboundary expansion for the union of determinantal hypertrees
Authors:
András Mészáros
Abstract:
We prove that for any large enough constant $k$, the union of $k$ independent $d$-dimensional determinantal hypertrees is a coboundary expander with high probability.
We prove that for any large enough constant $k$, the union of $k$ independent $d$-dimensional determinantal hypertrees is a coboundary expander with high probability.
△ Less
Submitted 12 December, 2023; v1 submitted 29 November, 2023;
originally announced November 2023.
-
Last passage percolation in a product-type random environment
Authors:
Yuri Bakhtin,
Konstantin Khanin,
András Mészáros,
Jeremy Voltz
Abstract:
We consider a last passage percolation model in dimension $1+1$ with potential given by the product of a spatial i.i.d. potential with symmetric bounded distribution and an independent i.i.d. in time sequence of signs. We assume that the density of the spatial potential near the edge of its support behaves as a power, with exponent $κ>-1$. We investigate the linear growth rate of the actions of op…
▽ More
We consider a last passage percolation model in dimension $1+1$ with potential given by the product of a spatial i.i.d. potential with symmetric bounded distribution and an independent i.i.d. in time sequence of signs. We assume that the density of the spatial potential near the edge of its support behaves as a power, with exponent $κ>-1$. We investigate the linear growth rate of the actions of optimal point-to-point lazy random walk paths as a function of the path slope and describe the structure of the resulting shape function. It has a corner at $0$ and, although its restriction to positive slopes cannot be linear, we prove that it has a flat edge near $0$ if $κ>0$.
For optimal point-to-line paths, we study their actions and locations of favorable edges that the paths tends to reach and stay at. Under an additional assumption on the time it takes for the optimal path to reach the favorable location, we prove that appropriately normalized actions converge to a limiting distribution that can be viewed as a counterpart of the Tracy-Widom law. Since the scaling exponent and the limiting distribution depend only on the parameter $κ$, our results provide a description of a new universality class.
△ Less
Submitted 12 October, 2023;
originally announced October 2023.
-
Global Well-Posedness of Displacement Monotone Degenerate Mean Field Games Master Equations
Authors:
Mohit Bansil,
Alpár R. Mészáros,
Chenchen Mou
Abstract:
In this manuscript we construct global in time classical solutions to mean field games master equations in the lack of idiosyncratic noise in the individual agents' dynamics. These include both deterministic models and dynamics driven solely by a Brownian common noise. We consider a general class of non-separable Hamiltonians and final data functions that are supposed to be displacement monotone.…
▽ More
In this manuscript we construct global in time classical solutions to mean field games master equations in the lack of idiosyncratic noise in the individual agents' dynamics. These include both deterministic models and dynamics driven solely by a Brownian common noise. We consider a general class of non-separable Hamiltonians and final data functions that are supposed to be displacement monotone. Our main results unify and generalize in particular some of the well-posedness results on displacement monotone master equations obtained recently by Gangbo--Mészáros and Gangbo--Mészáros--Mou--Zhang.
△ Less
Submitted 30 August, 2023;
originally announced August 2023.
-
Cohen-Lenstra distribution for sparse matrices with determinantal biasing
Authors:
András Mészáros
Abstract:
Let us consider the following matrix $B_n$. The columns of $B_n$ are indexed with $[n]=\{1,2,\dots,n\}$ and the rows are indexed with $[n]^3$. The row corresponding to $(x_1,x_2,x_3)\in [n]^3$ is given by $\sum_{i=1}^3 e_{x_i}$, where $e_1,e_2,\dots,e_n$ is the standard basis of $\mathbb{R}^{[n]}$. Let $A_n$ be random $n\times n$ submatrix of $B_n$, where the probability that we choose a submatrix…
▽ More
Let us consider the following matrix $B_n$. The columns of $B_n$ are indexed with $[n]=\{1,2,\dots,n\}$ and the rows are indexed with $[n]^3$. The row corresponding to $(x_1,x_2,x_3)\in [n]^3$ is given by $\sum_{i=1}^3 e_{x_i}$, where $e_1,e_2,\dots,e_n$ is the standard basis of $\mathbb{R}^{[n]}$. Let $A_n$ be random $n\times n$ submatrix of $B_n$, where the probability that we choose a submatrix $C$ is proportional to $|\det(C)|^2$.
Let $p\ge 5$ be a prime. We prove that the asymptotic distribution of the $p$-Sylow subgroup of the cokernel of $A_n$ is given by the Cohen-Lenstra heuristics.
Our result is motivated by the conjecture that the first homology group of a random two dimensional hypertree is also Cohen-Lenstra distributed.
△ Less
Submitted 10 July, 2023;
originally announced July 2023.
-
Learning Distributions over Trajectories for Human Behavior Prediction
Authors:
Anna Mészáros,
Julian F. Schumann,
Javier Alonso-Mora,
Arkady Zgonnikov,
Jens Kober
Abstract:
Predicting the future behavior of human road users is an important aspect for the development of risk-aware autonomous vehicles. While many models have been developed towards this end, effectively capturing and predicting the variability inherent to human behavior still remains an open challenge. This paper proposes TrajFlow - a new approach for probabilistic trajectory prediction based on Normali…
▽ More
Predicting the future behavior of human road users is an important aspect for the development of risk-aware autonomous vehicles. While many models have been developed towards this end, effectively capturing and predicting the variability inherent to human behavior still remains an open challenge. This paper proposes TrajFlow - a new approach for probabilistic trajectory prediction based on Normalizing Flows. We reformulate the problem of capturing distributions over trajectories into capturing distributions over abstracted trajectory features using an autoencoder, simplifying the learning task of the Normalizing Flows. TrajFlow outperforms state-of-the-art behavior prediction models in capturing full trajectory distributions in two synthetic benchmarks with known true distributions, and is competitive on the naturalistic datasets ETH/UCY, rounD, and nuScenes. Our results demonstrate the effectiveness of TrajFlow in probabilistic prediction of human behavior.
△ Less
Submitted 19 April, 2024; v1 submitted 11 April, 2023;
originally announced April 2023.
-
Eigenvectors of the square grid plus GUE
Authors:
András Mészáros,
Bálint Virág
Abstract:
Eigenvectors of the GUE-perturbed discrete torus with uniform boundary conditions retain some product structure for small perturbations but converge to discrete Gaussian waves for large perturbations. We determine where this phase transition happens.
Eigenvectors of the GUE-perturbed discrete torus with uniform boundary conditions retain some product structure for small perturbations but converge to discrete Gaussian waves for large perturbations. We determine where this phase transition happens.
△ Less
Submitted 14 November, 2023; v1 submitted 19 December, 2022;
originally announced December 2022.
-
Rethinking Sharpness-Aware Minimization as Variational Inference
Authors:
Szilvia Ujváry,
Zsigmond Telek,
Anna Kerekes,
Anna Mészáros,
Ferenc Huszár
Abstract:
Sharpness-aware minimization (SAM) aims to improve the generalisation of gradient-based learning by seeking out flat minima. In this work, we establish connections between SAM and Mean-Field Variational Inference (MFVI) of neural network parameters. We show that both these methods have interpretations as optimizing notions of flatness, and when using the reparametrisation trick, they both boil dow…
▽ More
Sharpness-aware minimization (SAM) aims to improve the generalisation of gradient-based learning by seeking out flat minima. In this work, we establish connections between SAM and Mean-Field Variational Inference (MFVI) of neural network parameters. We show that both these methods have interpretations as optimizing notions of flatness, and when using the reparametrisation trick, they both boil down to calculating the gradient at a perturbed version of the current mean parameter. This thinking motivates our study of algorithms that combine or interpolate between SAM and MFVI. We evaluate the proposed variational algorithms on several benchmark datasets, and compare their performance to variants of SAM. Taking a broader perspective, our work suggests that SAM-like updates can be used as a drop-in replacement for the reparametrisation trick.
△ Less
Submitted 19 October, 2022;
originally announced October 2022.
-
On monotonicity conditions for Mean Field Games
Authors:
P. Jameson Graber,
Alpár R. Mészáros
Abstract:
In this paper we propose two new monotonicity conditions that could serve as sufficient conditions for uniqueness of Nash equilibria in mean field games. In this study we aim for $unconditional\ uniqueness$ that is independent of the length of the time horizon, the regularity of the starting distribution of the agents, or the regularization effect of a non-degenerate idiosyncratic noise. Through a…
▽ More
In this paper we propose two new monotonicity conditions that could serve as sufficient conditions for uniqueness of Nash equilibria in mean field games. In this study we aim for $unconditional\ uniqueness$ that is independent of the length of the time horizon, the regularity of the starting distribution of the agents, or the regularization effect of a non-degenerate idiosyncratic noise. Through a rich class of simple examples we show that these new conditions are not only in dichotomy with each other, but also with the two widely studied monotonicity conditions in the literature, the Lasry-Lions monotonicity and displacement monotonicity conditions.
△ Less
Submitted 20 July, 2023; v1 submitted 5 October, 2022;
originally announced October 2022.
-
On Some Mean Field Games and Master Equations through the lens of conservation laws
Authors:
P. Jameson Graber,
Alpár R. Mészáros
Abstract:
In this manuscript we derive a new nonlinear transport equation written on the space of probability measures that allows to study a class of deterministic mean field games and master equations, where the interaction of the agents happens only at the terminal time. The point of view via this transport equation has two important consequences. First, this equation reveals a new monotonicity condition…
▽ More
In this manuscript we derive a new nonlinear transport equation written on the space of probability measures that allows to study a class of deterministic mean field games and master equations, where the interaction of the agents happens only at the terminal time. The point of view via this transport equation has two important consequences. First, this equation reveals a new monotonicity condition that is sufficient both for the uniqueness of MFG Nash equilibria and for the global in time well-posedness of master equations. Interestingly, this condition is in general in dichotomy with both the Lasry--Lions and displacement monotonicity conditions, studied so far in the literature. Second, in the absence of monotonicity, the conservative form of the transport equation can be used to define weak entropy solutions to the master equation. We construct several concrete examples to demonstrate that MFG Nash equilibria, whether or not they actually exist, may not be selected by the entropy solutions of the master equation.
△ Less
Submitted 21 March, 2024; v1 submitted 22 August, 2022;
originally announced August 2022.
-
Optimized numerical inverse Laplace transformation
Authors:
Illes Horvath,
Andras Meszaros,
Miklos Telek
Abstract:
Among the numerical inverse Laplace transformation (NILT) methods, those that belong to the Abate-Whitt framework (AWF) are considered to be the most efficient ones currently. It is a characteristic feature of the AWF NILT procedures that they are independent of the transform function and the time point of interest.
In this work we propose an NILT procedure that goes beyond this limitation and o…
▽ More
Among the numerical inverse Laplace transformation (NILT) methods, those that belong to the Abate-Whitt framework (AWF) are considered to be the most efficient ones currently. It is a characteristic feature of the AWF NILT procedures that they are independent of the transform function and the time point of interest.
In this work we propose an NILT procedure that goes beyond this limitation and optimize the accuracy of the NILT utilizing also the transform function and the time point of interest.
△ Less
Submitted 15 December, 2021;
originally announced December 2021.
-
A variational approach to first order kinetic Mean Field Games with local couplings
Authors:
Megan Griffin-Pickering,
Alpár R. Mészáros
Abstract:
First order kinetic mean field games formally describe the Nash equilibria of deterministic differential games where agents control their acceleration, asymptotically in the limit as the number of agents tends to infinity. The known results for the well-posedness theory of mean field games with control on the acceleration assume either that the running and final costs are regularizing functionals…
▽ More
First order kinetic mean field games formally describe the Nash equilibria of deterministic differential games where agents control their acceleration, asymptotically in the limit as the number of agents tends to infinity. The known results for the well-posedness theory of mean field games with control on the acceleration assume either that the running and final costs are regularizing functionals of the density variable, or the presence of noise, i.e. a second-order system. In this article we construct global in time weak solutions to a first order mean field games system involving kinetic transport operators, where the costs are local (hence non-regularizing) functions of the density variable with polynomial growth. We show the uniqueness of these solutions on the support of the agent density. This is achieved by characterizing solutions through two convex optimization problems in duality. As part of our approach, we develop tools for the analysis of mean field games on a non-compact domain by variational methods. We introduce a notion of `reachable set', built from the initial measure, that allows us to work with initial measures with or without compact support. In this way we are able to obtain crucial estimates on minimizing sequences for merely bounded and continuous initial measures. These are then carefully combined with $L^1$-type averaging lemmas from kinetic theory to obtain pre-compactness for the minimizing sequence. Finally, under stronger convexity and monotonicity assumptions on the data, we prove higher order Sobolev estimates of the solutions.
△ Less
Submitted 11 July, 2022; v1 submitted 6 December, 2021;
originally announced December 2021.
-
Depth Without the Magic: Inductive Bias of Natural Gradient Descent
Authors:
Anna Kerekes,
Anna Mészáros,
Ferenc Huszár
Abstract:
In gradient descent, changing how we parametrize the model can lead to drastically different optimization trajectories, giving rise to a surprising range of meaningful inductive biases: identifying sparse classifiers or reconstructing low-rank matrices without explicit regularization. This implicit regularization has been hypothesised to be a contributing factor to good generalization in deep lear…
▽ More
In gradient descent, changing how we parametrize the model can lead to drastically different optimization trajectories, giving rise to a surprising range of meaningful inductive biases: identifying sparse classifiers or reconstructing low-rank matrices without explicit regularization. This implicit regularization has been hypothesised to be a contributing factor to good generalization in deep learning. However, natural gradient descent is approximately invariant to reparameterization, it always follows the same trajectory and finds the same optimum. The question naturally arises: What happens if we eliminate the role of parameterization, which solution will be found, what new properties occur? We characterize the behaviour of natural gradient flow in deep linear networks for separable classification under logistic loss and deep matrix factorization. Some of our findings extend to nonlinear neural networks with sufficient but finite over-parametrization. We demonstrate that there exist learning problems where natural gradient descent fails to generalize, while gradient descent with the right architecture performs well.
△ Less
Submitted 22 November, 2021;
originally announced November 2021.
-
Learning to Pick at Non-Zero-Velocity from Interactive Demonstrations
Authors:
Anna Mészáros,
Giovanni Franzese,
Jens Kober
Abstract:
This work investigates how the intricate task of a continuous pick & place (P&P) motion may be learned from humans based on demonstrations and corrections. Due to the complexity of the task, these demonstrations are often slow and even slightly flawed, particularly at moments when multiple aspects (i.e., end-effector movement, orientation, and gripper width) have to be demonstrated at once. Rather…
▽ More
This work investigates how the intricate task of a continuous pick & place (P&P) motion may be learned from humans based on demonstrations and corrections. Due to the complexity of the task, these demonstrations are often slow and even slightly flawed, particularly at moments when multiple aspects (i.e., end-effector movement, orientation, and gripper width) have to be demonstrated at once. Rather than training a person to give better demonstrations, non-expert users are provided with the ability to interactively modify the dynamics of their initial demonstration through teleoperated corrective feedback. This in turn allows them to teach motions outside of their own physical capabilities. In the end, the goal is to obtain a faster but reliable execution of the task. The presented framework learns the desired movement dynamics based on the current Cartesian position with Gaussian Processes (GPs), resulting in a reactive, time-invariant policy. Using GPs also allows online interactive corrections and active disturbance rejection through epistemic uncertainty minimization. The experimental evaluation of the framework is carried out on a Franka-Emika Panda. Tests were performed to determine i) the framework's effectiveness in successfully learning how to quickly pick & place an object, ii) ease of policy correction to environmental changes (i.e., different object sizes and mass), and iii) the framework's usability for non-expert users.
△ Less
Submitted 11 April, 2022; v1 submitted 9 October, 2021;
originally announced October 2021.
-
Mean Field Games systems under displacement monotonicity
Authors:
Alpár R. Mészáros,
Chenchen Mou
Abstract:
In this note we prove the uniqueness of solutions to a class of Mean Field Games systems subject to possibly degenerate individual noise. Our results hold true for arbitrary long time horizons and for general non-separable Hamiltonians that satisfy a so-called $displacement\ monotonicity$ condition. This monotonicity condition that we propose for non-separable Hamiltonians is sharper and more gene…
▽ More
In this note we prove the uniqueness of solutions to a class of Mean Field Games systems subject to possibly degenerate individual noise. Our results hold true for arbitrary long time horizons and for general non-separable Hamiltonians that satisfy a so-called $displacement\ monotonicity$ condition. This monotonicity condition that we propose for non-separable Hamiltonians is sharper and more general than the one proposed in our earlier work written jointly with Gangbo and Zhang. The displacement monotonicity assumptions imposed on the data provide actually not only uniqueness, but also the existence and regularity of the solutions. Our analysis uses elementary arguments and does not rely on the well-posedness of the corresponding master equations.
△ Less
Submitted 22 August, 2023; v1 submitted 14 September, 2021;
originally announced September 2021.
-
Well-posedness of mean field games master equations involving non-separable local Hamiltonians
Authors:
David M. Ambrose,
Alpár R. Mészáros
Abstract:
In this paper we construct short time classical solutions to a class of master equations in the presence of non-degenerate individual noise arising in the theory of mean field games. The considered Hamiltonians are non-separable and $local$ functions of the measure variable, therefore the equation is restricted to absolutely continuous measures whose densities lie in suitable Sobolev spaces. Our r…
▽ More
In this paper we construct short time classical solutions to a class of master equations in the presence of non-degenerate individual noise arising in the theory of mean field games. The considered Hamiltonians are non-separable and $local$ functions of the measure variable, therefore the equation is restricted to absolutely continuous measures whose densities lie in suitable Sobolev spaces. Our results hold for smooth enough Hamiltonians, without any additional structural conditions as convexity or monotonicity.
△ Less
Submitted 15 June, 2022; v1 submitted 9 May, 2021;
originally announced May 2021.
-
ILoSA: Interactive Learning of Stiffness and Attractors
Authors:
Giovanni Franzese,
Anna Mészáros,
Luka Peternel,
Jens Kober
Abstract:
Teaching robots how to apply forces according to our preferences is still an open challenge that has to be tackled from multiple engineering perspectives. This paper studies how to learn variable impedance policies where both the Cartesian stiffness and the attractor can be learned from human demonstrations and corrections with a user-friendly interface. The presented framework, named ILoSA, uses…
▽ More
Teaching robots how to apply forces according to our preferences is still an open challenge that has to be tackled from multiple engineering perspectives. This paper studies how to learn variable impedance policies where both the Cartesian stiffness and the attractor can be learned from human demonstrations and corrections with a user-friendly interface. The presented framework, named ILoSA, uses Gaussian Processes for policy learning, identifying regions of uncertainty and allowing interactive corrections, stiffness modulation and active disturbance rejection. The experimental evaluation of the framework is carried out on a Franka-Emika Panda in four separate cases with unique force interaction properties: 1) pulling a plug wherein a sudden force discontinuity occurs upon successful removal of the plug, 2) pushing a box where a sustained force is required to keep the robot in motion, 3) wi** a whiteboard in which the force is applied perpendicular to the direction of movement, and 4) inserting a plug to verify the usability for precision-critical tasks in an experimental validation performed with non-expert users.
△ Less
Submitted 17 September, 2021; v1 submitted 4 March, 2021;
originally announced March 2021.
-
Mean Field Games Master Equations with Non-separable Hamiltonians and Displacement Monotonicity
Authors:
Wilfrid Gangbo,
Alpár R. Mészáros,
Chenchen Mou,
Jianfeng Zhang
Abstract:
In this manuscript, we propose a structural condition on non-separable Hamiltonians, which we term displacement monotonicity condition, to study second order mean field games master equations. A rate of dissipation of a bilinear form is brought to bear a global (in time) well-posedness theory, based on a--priori uniform Lipschitz estimates on the solution in the measure variable. Displacement mono…
▽ More
In this manuscript, we propose a structural condition on non-separable Hamiltonians, which we term displacement monotonicity condition, to study second order mean field games master equations. A rate of dissipation of a bilinear form is brought to bear a global (in time) well-posedness theory, based on a--priori uniform Lipschitz estimates on the solution in the measure variable. Displacement monotonicity being sometimes in dichotomy with the widely used Lasry-Lions monotonicity condition, the novelties of this work persist even when restricted to separable Hamiltonians.
△ Less
Submitted 1 April, 2022; v1 submitted 28 January, 2021;
originally announced January 2021.
-
The local weak limit of $k$-dimensional hypertrees
Authors:
András Mészáros
Abstract:
Let $\mathcal{C}(n,k)$ be the set of $k$-dimensional simplicial complexes $C$ over a fixed set of $n$ vertices such that:
(1) $C$ has a complete $k-1$-skeleton;
(2) $C$ has precisely ${{n-1}\choose {k}}$ $k$-faces;
(3) the homology group $H_{k-1}(C)$ is finite.
Consider the probability measure on $\mathcal{C}(n,k)$ where the probability of a simplicial complex $C$ is proportional to…
▽ More
Let $\mathcal{C}(n,k)$ be the set of $k$-dimensional simplicial complexes $C$ over a fixed set of $n$ vertices such that:
(1) $C$ has a complete $k-1$-skeleton;
(2) $C$ has precisely ${{n-1}\choose {k}}$ $k$-faces;
(3) the homology group $H_{k-1}(C)$ is finite.
Consider the probability measure on $\mathcal{C}(n,k)$ where the probability of a simplicial complex $C$ is proportional to $|H_{k-1}(C)|^2$. For any fixed $k$, we determine the local weak limit of these random simplicial complexes as $n$ tends to infinity.
This local weak limit turns out to be the same as the local weak limit of the $1$-out $k$-complexes investigated by Linial and Peled.
△ Less
Submitted 9 February, 2021; v1 submitted 27 January, 2021;
originally announced January 2021.
-
On the free energy density of factor models on biregular graphs
Authors:
András Mészáros
Abstract:
Let $h(0),h(1),\dots,h(k)$ be a symmetric concave sequence. For a $(d,k)$-biregular factor graph $G$ and $x\in \{0,1\}^V$, we define the Hamiltonian \[H_G(x)=\sum_{f\in F} h\left(\sum_{v\in \partial f} x_v\right),\] where $V$ is the set of variable nodes, $F$ is the set of factor nodes. We prove that if $(G_n)$ is a large girth sequence of $(d,k)$-biregular factor graphs, then the free energy dens…
▽ More
Let $h(0),h(1),\dots,h(k)$ be a symmetric concave sequence. For a $(d,k)$-biregular factor graph $G$ and $x\in \{0,1\}^V$, we define the Hamiltonian \[H_G(x)=\sum_{f\in F} h\left(\sum_{v\in \partial f} x_v\right),\] where $V$ is the set of variable nodes, $F$ is the set of factor nodes. We prove that if $(G_n)$ is a large girth sequence of $(d,k)$-biregular factor graphs, then the free energy density of $G_n$ converges. The limiting free energy density is given by the Bethe-approximation.
△ Less
Submitted 12 November, 2020;
originally announced November 2020.
-
Matchings on trees and the adjacency matrix: A determinantal viewpoint
Authors:
András Mészáros
Abstract:
Let $G$ be a finite tree. For any matching $M$ of $G$, let $U(M)$ be the set of vertices uncovered by $M$. Let $\mathcal{M}_G$ be a uniform random maximum size matching of $G$. In this paper, we analyze the structure of $U(\mathcal{M}_G)$. We first show that $U(\mathcal{M}_G)$ is a determinantal process. We also show that for most vertices of $G$, the process $U(\mathcal{M}_G)$ in a small neighbor…
▽ More
Let $G$ be a finite tree. For any matching $M$ of $G$, let $U(M)$ be the set of vertices uncovered by $M$. Let $\mathcal{M}_G$ be a uniform random maximum size matching of $G$. In this paper, we analyze the structure of $U(\mathcal{M}_G)$. We first show that $U(\mathcal{M}_G)$ is a determinantal process. We also show that for most vertices of $G$, the process $U(\mathcal{M}_G)$ in a small neighborhood of that vertex can be well approximated based on a somewhat larger neighborhood of the same vertex. Then we show that the normalized Shannon entropy of $U(\mathcal{M}_G)$ can be also well approximated using the local structure of $G$. In other words, in the realm of trees, the normalized Shannon entropy of $U(\mathcal{M}_G)$ -- that is, the normalized logarithm of the number of maximum size matchings of $G$ -- is a Benjamini-Schramm continuous parameter.
We show that $U(\mathcal{M}_G)$ is a determinantal process through establishing a new connection between $U(\mathcal{M}_G)$ and the adjacency matrix of $G$. This result sheds a new light on the well-known fact that on a tree, the number of vertices uncovered by a maximum size matching is equal to the nullity of the adjacency matrix.
Some of the proofs are based on the well established method of introducing a new perturbative parameter, which we call temperature, and then define the positive temperature analogue of $\mathcal{M}_G$, the so called monomer-dimer model, and let the temperature go to zero.
△ Less
Submitted 27 November, 2020; v1 submitted 8 November, 2020;
originally announced November 2020.
-
On nonexpansiveness of metric projection operators on Wasserstein spaces
Authors:
Anshul Adve,
Alpár Mészáros
Abstract:
In this paper we investigate properties of metric projections onto specific closed and geodesically convex proper subsets of Wasserstein spaces $(\mathcal{P}_p(\mathbf{R}^d),W_p).$ When $d=1$, as $(\mathcal{P}_2(\mathbf{R}),W_2)$ is isometrically isomorphic to a flat space with a Hilbertian structure, the corresponding projection operators are expected to be nonexpansive. We give a direct proof of…
▽ More
In this paper we investigate properties of metric projections onto specific closed and geodesically convex proper subsets of Wasserstein spaces $(\mathcal{P}_p(\mathbf{R}^d),W_p).$ When $d=1$, as $(\mathcal{P}_2(\mathbf{R}),W_2)$ is isometrically isomorphic to a flat space with a Hilbertian structure, the corresponding projection operators are expected to be nonexpansive. We give a direct proof of this fact, relying on intrinsic analysis, which also implies nonexpansiveness in certain special cases in higher dimensions. When $d>1$, we show the failure of this property in two regimes: when $p>1$ is either small enough or large enough. Finally, we prove some positive curvature properties of Wasserstein spaces $(\mathcal{P}_p(\mathbf{R}^d),W_p)$ when $d\ge 2$ and $p\in(1,+\infty)$ are arbitrary: we show that Wasserstein spaces are nowhere locally Busemann NPC spaces, and they nowhere locally satisfy the so-called projection criterion. As a corollary of the former, they have nonnegative upper Alexandrov curvature, in a precise sense that we define here. In our analysis a particular subset of probability measures having densities uniformly bounded above by a given constant plays a special role.
△ Less
Submitted 19 February, 2024; v1 submitted 2 September, 2020;
originally announced September 2020.
-
A BK inequality for random matchings
Authors:
András Mészáros
Abstract:
Let $G=(S,T,E)$ be a bipartite graph. For a matching $M$ of $G$, let $V(M)$ be the set of vertices covered by $M$, and let $B(M)$ be the symmetric difference of $V(M)$ and $S$. We prove that if $M$ is a uniform random matching of $G$, then $B(M)$ satisfies the BK inequality for increasing events.
Let $G=(S,T,E)$ be a bipartite graph. For a matching $M$ of $G$, let $V(M)$ be the set of vertices covered by $M$, and let $B(M)$ be the symmetric difference of $V(M)$ and $S$. We prove that if $M$ is a uniform random matching of $G$, then $B(M)$ satisfies the BK inequality for increasing events.
△ Less
Submitted 12 June, 2020;
originally announced June 2020.
-
Atoms of the matching measure
Authors:
Ferenc Bencs,
András Mészáros
Abstract:
We prove that the matching measure of an infinite vertex-transitive connected graph has no atoms. Generalizing the results of Salez, we show that for an ergodic non-amenable unimodular random rooted graph with uniformly bounded degrees, the matching measure has only finitely many atoms. Ku and Chen proved the analogue of the Gallai-Edmonds structure theorem for non-zero roots of the matching polyn…
▽ More
We prove that the matching measure of an infinite vertex-transitive connected graph has no atoms. Generalizing the results of Salez, we show that for an ergodic non-amenable unimodular random rooted graph with uniformly bounded degrees, the matching measure has only finitely many atoms. Ku and Chen proved the analogue of the Gallai-Edmonds structure theorem for non-zero roots of the matching polynomial for finite graphs. We extend their results for infinite graphs. We also show that the corresponding Gallai-Edmonds decomposition is compatible with the zero temperature monomer-dimer model.
△ Less
Submitted 25 May, 2020;
originally announced May 2020.
-
Global well-posedness of Master equations for deterministic displacement convex potential mean field games
Authors:
Wilfrid Gangbo,
Alpár R. Mészáros
Abstract:
This manuscript constructs global in time solutions to the $master\ equations$ for potential Mean Field Games. The study concerns a class of Lagrangians and initial data functions, which are $displacement\ convex$ and so, it may be in dichotomy with the class of so--called $monotone$ functions, widely considered in the literature. We construct solutions to both the scalar and vectorial master equa…
▽ More
This manuscript constructs global in time solutions to the $master\ equations$ for potential Mean Field Games. The study concerns a class of Lagrangians and initial data functions, which are $displacement\ convex$ and so, it may be in dichotomy with the class of so--called $monotone$ functions, widely considered in the literature. We construct solutions to both the scalar and vectorial master equations in potential Mean Field Games, when the underlying space is the whole space $\mathbb{R}^d$ and so, it is not compact.
△ Less
Submitted 18 October, 2021; v1 submitted 3 April, 2020;
originally announced April 2020.
-
Gamma-ray bursts: A brief survey of the diversity
Authors:
Attila Mészáros,
Jakub Řípa
Abstract:
The separation of the gamma-ray bursts (GRBs) into short/hard and long/soft subclasses, respectively, is well supported both theoretically and observationally. The long ones are coupled to supernovae type Ib/Ic - the short ones are connected to the merging of two neutron stars, where one or even both neutron stars can be substituted by black holes. These short GRBs - as merging binaries - can also…
▽ More
The separation of the gamma-ray bursts (GRBs) into short/hard and long/soft subclasses, respectively, is well supported both theoretically and observationally. The long ones are coupled to supernovae type Ib/Ic - the short ones are connected to the merging of two neutron stars, where one or even both neutron stars can be substituted by black holes. These short GRBs - as merging binaries - can also serve as the sources of gravitation waves, and are observable as the recently detected macronovae. Since 1998 there are several statistical studies suggesting the existence of more than two subgroups. There can be a subgroup having an intermediate durations; there can be a subgroup with ultra-long durations; the long/soft subgroup itself can be divided into two subclasses with respect to the luminosity of GRBs. The authors with other collaborators provided several statistical studies in this topic. This field of the GRB-diversity is briefly surveyed in this contribution.
△ Less
Submitted 3 January, 2020;
originally announced January 2020.
-
The analogy of K-correction in the topic of gamma-ray bursts
Authors:
Levente Borvák,
Attila Mészáros,
Jakub Řípa
Abstract:
It is well-known that there are two types of gamma-ray bursts (GRBs): short/hard and long/soft ones, respectively. The long GRBs are coupled to supernovae, but the short ones are associated with the so called macronovae (also known as kilonovae), which can serve as the sources of gravitational waves as well. The kilonovae can arise from the merging of two neutron-stars. The neutron stars can be su…
▽ More
It is well-known that there are two types of gamma-ray bursts (GRBs): short/hard and long/soft ones, respectively. The long GRBs are coupled to supernovae, but the short ones are associated with the so called macronovae (also known as kilonovae), which can serve as the sources of gravitational waves as well. The kilonovae can arise from the merging of two neutron-stars. The neutron stars can be substituded by more massive black holes as well. Hence, the topic of gamma-ray bursts (mainly the topic of short ones) and the topic of massive binaries, are strongly connected.
In this contribution, the redshifts of GRBs are studied. The surprising result - namely that the apparently fainter GRBs can be in average at smaller distances - is discussed again. In essence, the results of Mészáros et al. (2011) are studied again using newer samples of GRBs. The former result is confirmed by the newer data.
△ Less
Submitted 3 January, 2020;
originally announced January 2020.
-
An Oppositeness in the Cosmology: Distribution of the Gamma-Ray Bursts and the Cosmological Principle
Authors:
Attila Meszaros
Abstract:
The Cosmological Principle is the assumption that the universe is spatially homogeneous and isotropic in the large-scale average. In year 1998 the author, together with his two colleagues, has shown that the BATSE's short gamma-ray bursts are not distributed isotropically on the sky. This claim was then followed by other papers confirming both the existence of anisotropies in the angular distribut…
▽ More
The Cosmological Principle is the assumption that the universe is spatially homogeneous and isotropic in the large-scale average. In year 1998 the author, together with his two colleagues, has shown that the BATSE's short gamma-ray bursts are not distributed isotropically on the sky. This claim was then followed by other papers confirming both the existence of anisotropies in the angular distribution of bursts and the existence of huge Gpc structures in the spatial distribution. These observational facts are in contradiction with the Cosmological Principle, because the large scale averaging hardly can be provided. The aim of this contribution is to survey these publications.
△ Less
Submitted 16 December, 2019;
originally announced December 2019.
-
Degenerate nonlinear parabolic equations with discontinuous diffusion coefficients
Authors:
Dohyun Kwon,
Alpár Richárd Mészáros
Abstract:
This paper is devoted to the study of some nonlinear parabolic equations with discontinuous diffusion intensities. Such problems appear naturally in physical and biological models. Our analysis is based on variational techniques and in particular on gradient flows in the space of probability measures equipped with the distance arising in the Monge-Kantorovich optimal transport problem. The associa…
▽ More
This paper is devoted to the study of some nonlinear parabolic equations with discontinuous diffusion intensities. Such problems appear naturally in physical and biological models. Our analysis is based on variational techniques and in particular on gradient flows in the space of probability measures equipped with the distance arising in the Monge-Kantorovich optimal transport problem. The associated internal energy functionals in general fail to be differentiable, therefore classical results do not apply directly in our setting. We study the combination of both linear and porous medium type diffusions and we show the existence and uniqueness of the solutions in the sense of distributions in suitable Sobolev spaces. Our notion of solution allows us to give a fine characterization of the emerging critical regions, observed previously in numerical experiments. A link to a three phase free boundary problem is also pointed out.
△ Less
Submitted 7 February, 2021; v1 submitted 25 July, 2019;
originally announced July 2019.
-
Limiting entropy of determinantal processes
Authors:
András Mészáros
Abstract:
We extend Lyons's tree entropy theorem to general determinantal measures. As a byproduct we show that the sofic entropy of an invariant determinantal measure does not depend on the chosen sofic approximation.
We extend Lyons's tree entropy theorem to general determinantal measures. As a byproduct we show that the sofic entropy of an invariant determinantal measure does not depend on the chosen sofic approximation.
△ Less
Submitted 25 May, 2020; v1 submitted 27 May, 2019;
originally announced May 2019.
-
Weak solutions to the Muskat problem with surface tension via optimal transport
Authors:
Matt Jacobs,
Inwon Kim,
Alpár R. Mészáros
Abstract:
Inspired by recent works on the threshold dynamics scheme for multi-phase mean curvature flow (by Esedoglu-Otto and Laux-Otto), we introduce a novel framework to approximate solutions of the Muskat problem with surface tension. Our approach is based on interpreting the Muskat problem as a gradient flow in a product Wasserstein space. This perspective allows us to construct weak solutions via a min…
▽ More
Inspired by recent works on the threshold dynamics scheme for multi-phase mean curvature flow (by Esedoglu-Otto and Laux-Otto), we introduce a novel framework to approximate solutions of the Muskat problem with surface tension. Our approach is based on interpreting the Muskat problem as a gradient flow in a product Wasserstein space. This perspective allows us to construct weak solutions via a minimizing movements scheme. Rather than working directly with the singular surface tension force, we instead relax the perimeter functional with the heat content energy approximation of Esedoglu-Otto. The heat content energy allows us to show the convergence of the associated minimizing movement scheme in the Wasserstein space, and makes the scheme far more tractable for numerical simulations. Under a typical energy convergence assumption, we show that our scheme converges to weak solutions of the Muskat problem with surface tension. We then conclude the paper with a discussion on some numerical experiments and on equilibrium configurations.
△ Less
Submitted 15 September, 2020; v1 submitted 13 May, 2019;
originally announced May 2019.
-
Multi timescale bandwidth profile and its application for burst-aware fairness
Authors:
Szilveszter Nádas,
Balázs Varga,
Illés Horváth,
András Mészáros,
Miklós Telek
Abstract:
We propose a resource sharing scheme that takes into account the traffic history over several predefined time scales and provides fair resource sharing considering the traffic history. Our concept builds on a simplified version of core-stateless resource sharing, where we only use a few Drop Precedences (DPs). For packet marking we introduce Multi timescale bandwidth profile. Additionally, we prov…
▽ More
We propose a resource sharing scheme that takes into account the traffic history over several predefined time scales and provides fair resource sharing considering the traffic history. Our concept builds on a simplified version of core-stateless resource sharing, where we only use a few Drop Precedences (DPs). For packet marking we introduce Multi timescale bandwidth profile. Additionally, we provide basic dimensioning concepts for the proposed schema and present its simulation based performance analysis.
△ Less
Submitted 19 March, 2019;
originally announced March 2019.
-
The planning problem in Mean Field Games as regularized mass transport
Authors:
P. Jameson Graber,
Alpár R. Mészáros,
Francisco J. Silva,
Daniela Tonon
Abstract:
In this paper, using variational approaches, we investigate the first order planning problem arising in the theory of mean field games. We show the existence and uniqueness of weak solutions of the problem in the case of a large class of Hamiltonians with arbitrary superlinear order of growth at infinity and local coupling functions. We require the initial and final measures to be merely summable.…
▽ More
In this paper, using variational approaches, we investigate the first order planning problem arising in the theory of mean field games. We show the existence and uniqueness of weak solutions of the problem in the case of a large class of Hamiltonians with arbitrary superlinear order of growth at infinity and local coupling functions. We require the initial and final measures to be merely summable. At the same time (relying on the techniques developed recently by Graber and Mészáros), under stronger monotonicity and convexity conditions on the data, we obtain Sobolev estimates on the solutions of the planning problem both for space and time derivatives.
△ Less
Submitted 18 May, 2019; v1 submitted 6 November, 2018;
originally announced November 2018.
-
The distribution of sandpile groups of random regular graphs
Authors:
András Mészáros
Abstract:
We study the distribution of the sandpile group of random d-regular graphs. For the directed model, we prove that it follows the Cohen-Lenstra heuristics, that is, the limiting probability that the $p$-Sylow subgroup of the sandpile group is a given $p$-group $P$, is proportional to $|Aut(P)|^{-1}$. For finitely many primes, these events get independent in the limit. Similar results hold for undir…
▽ More
We study the distribution of the sandpile group of random d-regular graphs. For the directed model, we prove that it follows the Cohen-Lenstra heuristics, that is, the limiting probability that the $p$-Sylow subgroup of the sandpile group is a given $p$-group $P$, is proportional to $|Aut(P)|^{-1}$. For finitely many primes, these events get independent in the limit. Similar results hold for undirected random regular graphs, where for odd primes the limiting distributions are the ones given by Clancy, Leake and Payne.
This answers an open question of Frieze and Vu whether the adjacency matrix of a random regular graph is invertible with high probability. Note that for directed graphs this was recently proved by Huang. It also gives an alternate proof of a theorem of Backhausz and Szegedy.
△ Less
Submitted 28 January, 2020; v1 submitted 10 June, 2018;
originally announced June 2018.
-
Sobolev regularity for first order Mean Field Games
Authors:
P. Jameson Graber,
Alpár R. Mészáros
Abstract:
In this paper we obtain Sobolev estimates for weak solutions of first oder variational Mean Field Game systems with coupling terms that are local function of the density variable. Under some coercivity condition on the coupling, we obtain first order Sobolev estimates for the density variable, while under similar coercivity condition on the Hamiltonian we obtain second order Sobolev estimates for…
▽ More
In this paper we obtain Sobolev estimates for weak solutions of first oder variational Mean Field Game systems with coupling terms that are local function of the density variable. Under some coercivity condition on the coupling, we obtain first order Sobolev estimates for the density variable, while under similar coercivity condition on the Hamiltonian we obtain second order Sobolev estimates for the value function. These results are valid both for stationary and time-dependent problems. In the latter case the estimates are fully global in time, thus we resolve a question which was left open in [PS17]. Our methods apply to a large class of Hamiltonians and coupling functions.
△ Less
Submitted 23 January, 2018; v1 submitted 21 August, 2017;
originally announced August 2017.
-
On nonlinear cross-diffusion systems: an optimal transport approach
Authors:
Inwon Kim,
Alpár R. Mészáros
Abstract:
We study a nonlinear, degenerate cross-diffusion model which involves two densities with two different drift velocities. A general framework is introduced based on its gradient flow structure in Wasserstein space to derive a notion of discrete-time solutions. Its continuum limit, due to the possible mixing of the densities, only solves a weaker version of the original system. In one space dimensio…
▽ More
We study a nonlinear, degenerate cross-diffusion model which involves two densities with two different drift velocities. A general framework is introduced based on its gradient flow structure in Wasserstein space to derive a notion of discrete-time solutions. Its continuum limit, due to the possible mixing of the densities, only solves a weaker version of the original system. In one space dimension, where the densities are guaranteed to be segregated, a stable interface appears between the two densities, and a stronger convergence result, in particular derivation of a standard weak solution to the system, is available. We also study the incompressible limit of the system, which addresses transport under a height constraint on the total density. In one space dimension we show that the problem leads to a two-phase Hele-Shaw type flow.
△ Less
Submitted 17 March, 2018; v1 submitted 6 May, 2017;
originally announced May 2017.
-
On the variational formulation of some stationary second order mean field games systems
Authors:
Alpár Richárd Mészáros,
Francisco J. Silva
Abstract:
We consider the variational approach to prove the existence of solutions of second order stationary Mean Field Games on a bounded domain $Ω\subseteq \mathbb{R}^{d}$, with Neumann boundary conditions, and with and without density constraints. We consider Hamiltonians which growth as $|\cdot|^{q'}$, where $q'=q/(q-1)$ and $q>d$. Despite this restriction, our approach allows us to prove the existenc…
▽ More
We consider the variational approach to prove the existence of solutions of second order stationary Mean Field Games on a bounded domain $Ω\subseteq \mathbb{R}^{d}$, with Neumann boundary conditions, and with and without density constraints. We consider Hamiltonians which growth as $|\cdot|^{q'}$, where $q'=q/(q-1)$ and $q>d$. Despite this restriction, our approach allows us to prove the existence of solutions in the case of rather general coupling terms. When density constraints are taken into account, our results improve those in \cite{MesSil}. Furthermore, our approach can be used to obtain solutions of systems with multiple populations.
△ Less
Submitted 18 April, 2017; v1 submitted 7 April, 2017;
originally announced April 2017.
-
Adinkras From Ordered Quartets of BC${}_4$ Coxeter Group Elements and Regarding 1,358,954,496 Matrix Elements of the Gadget
Authors:
S. James Gates Jr.,
Forrest Guyton,
Siddhartha Harmalkar,
David S. Kessler,
Vadim Korotkikh,
Victor A. Meszaros
Abstract:
We examine values of the Adinkra Holoraumy-induced Gadget representation space metric over all possible four-color, four-open node, and four-closed node adinkras. Of the 1,358,954,496 gadget matrix elements, only 226,492,416 are non-vanishing and take on one of three values: $-1/3$, $1/3$, or $1$ and thus a subspace isomorphic to a description of a body-centered tetrahedral molecule emerges.
We examine values of the Adinkra Holoraumy-induced Gadget representation space metric over all possible four-color, four-open node, and four-closed node adinkras. Of the 1,358,954,496 gadget matrix elements, only 226,492,416 are non-vanishing and take on one of three values: $-1/3$, $1/3$, or $1$ and thus a subspace isomorphic to a description of a body-centered tetrahedral molecule emerges.
△ Less
Submitted 24 May, 2017; v1 submitted 1 January, 2017;
originally announced January 2017.
-
On the Connection of Gamma-Ray Bursts and X-Ray Flashes in the BATSE and RHESSI Databases
Authors:
Jakub Řípa,
Attila Mészáros
Abstract:
Classification of gamma-ray bursts (GRBs) into groups has been intensively studied by various statistical tests in previous years. It has been suggested that there was a distinct group of GRBs, beyond the long and short ones, with intermediate durations. However, such a group is not securely confirmed yet. Strangely, concerning the spectral hardness, the observations from the Swift and RHESSI sate…
▽ More
Classification of gamma-ray bursts (GRBs) into groups has been intensively studied by various statistical tests in previous years. It has been suggested that there was a distinct group of GRBs, beyond the long and short ones, with intermediate durations. However, such a group is not securely confirmed yet. Strangely, concerning the spectral hardness, the observations from the Swift and RHESSI satellites give different results. For the Swift/BAT database it is found that the intermediate-duration bursts might well be related to so-called X-ray flashes (XRFs). On the other hand, for the RHESSI dataset the intermediate-duration bursts seem to be spectrally too hard to be given by XRFs. The connection of the intermediate-duration bursts and XRFs for the BATSE database is not clear as well. The purpose of this article is to check the relation between XRFs and GRBs for the BATSE and RHESSI databases, respectively. We use an empirical definition of XRFs introduced by other authors earlier. For the RHESSI database we also use a transformation between the detected counts and the fluences based on the simulated detector response function. The purpose is to compare the hardnesses of GRBs with the definition of XRFs. There is a 1.3-4.2 % fraction of XRFs in the whole BATSE database. The vast majority of the BATSE short bursts are not XRFs because only 0.7-5.7 % of the short bursts can be given by XRFs. However, there is a large uncertainty in the fraction of XRFs among the intermediate-duration bursts. The fraction of 1-85 % of the BATSE intermediate-duration bursts can be related to XRFs. For the long bursts this fraction is between 1.0 % and 3.4 %. The uncertainties in these fractions are large, however it can be claimed that all BATSE intermediate-duration bursts cannot be given by XRFs. At least 79 % of RHESSI short bursts, at least 53 % of RHESSI intermediate-duration bursts, and at least...
△ Less
Submitted 25 October, 2016;
originally announced October 2016.
-
What is the Astrophysical Meaning of the Intermediate Subgroup of GRBs?
Authors:
J. Ripa,
A. Meszaros
Abstract:
Published articles concerning the intermediate (third) subgroup of GRBs are surveyed. From a statistical perspective this subgroup may exist, however its significance depends on which data set is used. Its astrophysical meaning is unclear because the occurrence of this subgroup can also be an artificial selection effect. Hence, GRBs from this subgroup need not be given by a physically different ph…
▽ More
Published articles concerning the intermediate (third) subgroup of GRBs are surveyed. From a statistical perspective this subgroup may exist, however its significance depends on which data set is used. Its astrophysical meaning is unclear because the occurrence of this subgroup can also be an artificial selection effect. Hence, GRBs from this subgroup need not be given by a physically different phenomenon. The aim of this contribution is to search for the answer to the question in the title.
△ Less
Submitted 21 July, 2015;
originally announced July 2015.
-
Uniqueness issues for evolution equations with density constraints
Authors:
Simone Di Marino,
Alpár Richárd Mészáros
Abstract:
In this paper we present some basic uniqueness results for evolutive equations under density constraints. First, we develop a rigorous proof of a well-known result (among specialists) in the case where the spontaneous velocity field satisfies a monotonicity assumption: we prove the uniqueness of a solution for first order systems modeling crowd motion with hard congestion effects, introduced recen…
▽ More
In this paper we present some basic uniqueness results for evolutive equations under density constraints. First, we develop a rigorous proof of a well-known result (among specialists) in the case where the spontaneous velocity field satisfies a monotonicity assumption: we prove the uniqueness of a solution for first order systems modeling crowd motion with hard congestion effects, introduced recently by \emph{Maury et al.} The monotonicity of the velocity field implies that the $2-$Wasserstein distance along two solutions is $λ$-contractive, which in particular implies uniqueness. In the case of diffusive models, we prove the uniqueness of a solution passing through the dual equation, where we use some well-known parabolic estimates to conclude an $L^1-$contraction property. In this case, by the regularization effect of the non-degenerate diffusion, the result follows even if the given velocity field is only $L^\infty$ as in the standard Fokker-Planck equation.
△ Less
Submitted 18 April, 2017; v1 submitted 10 July, 2015;
originally announced July 2015.
-
First order Mean Field Games with density constraints: Pressure equals Price
Authors:
Pierre Cardaliaguet,
Alpár Richárd Mészáros,
Filippo Santambrogio
Abstract:
In this paper we study Mean Field Game systems under density constraints as optimality conditions of two optimization problems in duality. A weak solution of the system contains an extra term, an additional price imposed on the saturated zones. We show that this price corresponds to the pressure field from the models of incompressible Euler's equations {à} la Brenier. By this observation we manage…
▽ More
In this paper we study Mean Field Game systems under density constraints as optimality conditions of two optimization problems in duality. A weak solution of the system contains an extra term, an additional price imposed on the saturated zones. We show that this price corresponds to the pressure field from the models of incompressible Euler's equations {à} la Brenier. By this observation we manage to obtain a minimal regularity, which allows to write optimality conditions at the level of single agent trajectories and to define a weak notion of Nash equilibrium for our model.
△ Less
Submitted 8 December, 2016; v1 submitted 8 July, 2015;
originally announced July 2015.
