A dynamic programming approach to the Parisi functional

DOI10.1090/PROC/12968MaRDI QIDQ2802130zbMATH OpenOpenAlexFDO

Authors Aukosh Jagannath, Ian Tobasco

Publication date 25 April 2016

Published in Proceedings of the American Mathematical Society (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1502.04398

zbMATH Keywords

dynamic programmingParisi formulaSherrington-Kirkpatrick model

Mathematics Subject Classification ID

Applications of optimal control and differential games (49N90) Interacting random processes; statistical mechanics type models; percolation theory (60K35) Disordered systems (random Ising models, random Schrödinger operators, etc.) in equilibrium statistical mechanics (82B44) Statistical mechanics of random media, disordered materials (including liquid crystals and spin glasses) (82D30)

Abstract: G.Parisi predicted an important variational formula for the thermodynamic limit of the intensive free energy for a class of mean field spin glasses. In this paper, we present an elementary approach to the study of the Parisi functional using stochastic dynamic programing and semi-linear PDE. We give a derivation of important properties of the Parisi PDE avoiding the use of Ruelle Probability Cascades and Cole-Hopf transformations. As an application, we give a simple proof of the strict convexity of the Parisi functional, which was recently proved by Auffinger and Chen in [2].

Review text

The paper deals with the Parisi functional and its main properties. The Parisi functional is determined by a solution of the Parisi partial differential equation depending on a measure. In the case that the measure has many atoms, the main properties of a solution of the Parisi PDE are well known. In case of the general measure, the question of uniqueness of a solution of the Parisi PDE was open. Using standard arguments from semi-linear parabolic PDEs, the authors prove the existence, uniqueness and regularity of the solution. In this paper, the strict convexity of the Parisi functional is also proved as an application of the properties of the Parisi PDE.

Recommendations

• Partial results on the convexity of the Parisi functional with PDE approach
• On properties of Parisi measures
• A question about the Parisi functional
• Parisi measures
• The Parisi formula has a unique minimizer

Cites work

• A question about the Parisi functional
• A variational representation for certain functionals of Brownian motion
• Exit probabilities and optimal stochastic control
• scientific article; zbMATH DE number 3505708 (Why is no real title available?)
• scientific article; zbMATH DE number 1273988 (Why is no real title available?)
• scientific article; zbMATH DE number 1751767 (Why is no real title available?)
• scientific article; zbMATH DE number 826154 (Why is no real title available?)
• scientific article; zbMATH DE number 5681750 (Why is no real title available?)
• Mean Field Models for Spin Glasses
• Mean field models for spin glasses. Volume I: Basic examples.
• Multidimensional diffusion processes.
• On properties of Parisi measures
• Parisi measures
• Partial differential equations. 4th ed
• Partial results on the convexity of the Parisi functional with PDE approach
• The Aizenman-Sims-Starr and Guerra's schemes for the SK model with multidimensional spins
• The Parisi formula
• The Parisi formula for mixed \(p\)-spin models
• The Parisi formula has a unique minimizer
• The Sherrington-Kirkpatrick model

Cited in

(48)

• On the energy landscape of the mixed even \(p\)-spin model
• Disorder chaos in some diluted spin Glass models
• Spectral gap estimates in mean field spin glasses
• On the TAP free energy in the mixed p-spin models
• On spin distributions for generic p-spin models
• Localization of directed polymers with general reference walk
• On the unbalanced cut problem and the generalized Sherrington-Kirkpatrick model
• The overlap gap property in principal submatrix recovery
• Optimization of mean-field spin glasses
• On the Almeida-Thouless transition line in the Sherrington-Kirkpatrick model with centered Gaussian external field
• Free energy in multi-species mixed \(p\)-spin spherical models
• The \(\ell^p\)-Gaussian-Grothendieck problem with vector spins
• Proof of the satisfiability conjecture for large \(k\)
• Extending the Parisi formula along a Hamilton-Jacobi equation
• The generalized TAP free energy. II
• The overlap gap property and approximate message passing algorithms for \(p\)-spin models
• Phase transition in random tensors with multiple independent spikes
• Phase transition in the spiked random tensor with Rademacher prior
• Fluctuations of the free energy in the mixed \(p\)-spin models with external field
• Parisi measures
• Suboptimality of local algorithms for a class of max-cut problems
• A question about the Parisi functional
• Temperature chaos in some spherical mixed p-spin models
• Algorithmic pure states for the negative spherical perceptron
• On the Marchenko–Pastur law in analog bipartite spin-glasses
• Chaos in temperature in generic \(2p\)-spin models
• The Parisi formula is a Hamilton-Jacobi equation in Wasserstein space
• Optimization of the Sherrington--Kirkpatrick Hamiltonian
• Partial results on the convexity of the Parisi functional with PDE approach
• Some properties of the phase diagram for mixed p-spin glasses
• Free energy subadditivity for symmetric random Hamiltonians
• Generalized TAP Free Energy
• Local algorithms for maximum cut and minimum bisection on locally treelike regular graphs of large degree
• Ultrametricity in spin glasses
• A Friendly Tutorial on Mean-Field Spin Glass Techniques for Non-Physicists
• Optimizing mean field spin glasses with external field
• Parisi formula for balanced Potts spin glass
• On Parisi measures of Potts spin glasses with correction
• Tight Lipschitz hardness for optimizing mean field spin glasses
• Un-inverting the Parisi formula
• Parisi PDE and convexity for vector spins
• Hessian spectrum at the global minimum of the spherical pure-like mixed p-spin glasses
• Optimization of the Sherrington-Kirkpatrick Hamiltonian
• The Legendre structure of the Parisi formula
• A CLuP algorithm to practically achieve 0.76 SK-model ground state free energy
• On the phase diagram of the multiscale mean-field spin-glass
• Pseudo-maximum likelihood theory for high-dimensional rank one inference
• Spin glass computations and Ruelle's probability cascades

This page was built for publication: A dynamic programming approach to the Parisi functional

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2802130)