Phase transition in the spiked random tensor with Rademacher prior

DOI10.1214/18-AOS1763MaRDI QIDQ2328058zbMATH OpenFDO

Authors Wei-Kuo Chen

Publication date 9 October 2019

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1712.01777, https://projecteuclid.org/euclid.aos/1564797862

zbMATH Keywords

BBP transitionParisi formulareplica symmetry breakingsignal detectionspiked tensorspin glass

Mathematics Subject Classification ID

Interacting random processes; statistical mechanics type models; percolation theory (60K35) Symmetry breaking in quantum theory (81R40) Phase transitions (general) in equilibrium statistical mechanics (82B26) Critical phenomena in equilibrium statistical mechanics (82B27) 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: We consider the problem of detecting a deformation from a symmetric Gaussian random $p$-tensor $(pgeq3)$ with a rank-one spike sampled from the Rademacher prior. Recently in Lesieur et al. (2017), it was proved that there exists a critical threshold $$ so that when the signal-to-noise ratio exceeds $$, one can distinguish the spiked and unspiked tensors and weakly recover the prior via the minimal mean-square-error method. On the other side, Perry, Wein, and Bandeira (2017) proved that there exists a $$ such that any statistical hypothesis test can not distinguish these two tensors, in the sense that their total variation distance asymptotically vanishes, when the signa-to-noise ratio is less than $$. In this work, we show that $$ is indeed the critical threshold that strictly separates the distinguishability and indistinguishability between the two tensors under the total variation distance. Our approach is based on a subtle analysis of the high temperature behavior of the pure $p$-spin model with Ising spin, arising initially from the field of spin glasses. In particular, we identify the signal-to-noise criticality $$ as the critical temperature, distinguishing the high and low temperature behavior, of the Ising pure $p$-spin mean-field spin glass model.

Recommendations

• Phase transition in random tensors with multiple independent spikes
• Statistical limits of spiked tensor models
• Statistical thresholds for tensor PCA
• The landscape of the spiked tensor model
• Algorithmic thresholds for tensor PCA

Cites work

• A dynamic programming approach to the Parisi functional
• Asymptotic mutual information for the balanced binary stochastic block model
• Asymptotic power of sphericity tests for high-dimensional data
• Chaos in the mixed even-spin models
• Constrained low-rank matrix estimation: phase transitions, approximate message passing and applications
• Exact solution of the gauge symmetric \(p\)-spin glass model on a complete graph
• Fluctuations of the free energy in the REM and the \(p\)-spin SK models
• Free energy in the mixed p-spin models with vector spins
• Fundamental limits of symmetric low-rank matrix estimation
• scientific article; zbMATH DE number 1273988 (Why is no real title available?)
• Mean Field Models for Spin Glasses
• Mean field models for spin glasses. Volume I: Basic examples.
• Mutual Information and Optimality of Approximate Message-Passing in Random Linear Estimation
• Non-Negative Principal Component Analysis: Message Passing Algorithms and Sharp Asymptotics
• On the Limitation of Spectral Methods: From the Gaussian Hidden Clique Problem to Rank One Perturbations of Gaussian Tensors
• Optimality and sub-optimality of PCA. I: Spiked random matrix models
• Phase transition in the spiked random tensor with Rademacher prior
• Phase transition of the largest eigenvalue for nonnull complex sample covariance matrices
• Some rigorous results on the Sherrington-Kirkpatrick spin glass model.
• State evolution for general approximate message passing algorithms, with applications to spatial coupling
• The p-spin interaction model with external field
• The adaptive interpolation method: a simple scheme to prove replica formulas in Bayesian inference
• The Dynamics of Message Passing on Dense Graphs, with Applications to Compressed Sensing
• The eigenvalues and eigenvectors of finite, low rank perturbations of large random matrices
• The free energy in a multi-species Sherrington-Kirkpatrick model
• The largest eigenvalue of small rank perturbations of Hermitian random matrices
• The largest eigenvalues of finite rank deformation of large Wigner matrices: Convergence and nonuniversality of the fluctuations
• The Parisi formula
• The Parisi formula for mixed \(p\)-spin models
• The Parisi formula has a unique minimizer
• The Parisi ultrametricity conjecture
• The Sherrington-Kirkpatrick model
• Variational representations for the Parisi functional and the two-dimensional Guerra-Talagrand bound

Cited in

(28)

• Deep Boltzmann machines: rigorous results at arbitrary depth
• Sparse random tensors: concentration, regularization and applications
• An optimal statistical and computational framework for generalized tensor estimation
• Inference for low-rank tensors -- no need to debias
• Free energy in multi-species mixed \(p\)-spin spherical models
• Strong replica symmetry in high-dimensional optimal Bayesian inference
• Statistical limits of spiked tensor models
• Phase transition in random tensors with multiple independent spikes
• Phase transition in the spiked random tensor with Rademacher prior
• Statistical thresholds for tensor PCA
• Order of fluctuations of the free energy in the SK model at critical temperature
• Sharp complexity asymptotics and topological trivialization for the (p, k) spiked tensor model
• Free energy subadditivity for symmetric random Hamiltonians
• Marchenko–Pastur law with relaxed independence conditions
• The free energy of spherical pure p-spin models: computation from the TAP approach
• Ultrametricity in spin glasses
• A Friendly Tutorial on Mean-Field Spin Glass Techniques for Non-Physicists
• Free energy upper bound for mean-field vector spin glasses
• On the second moment method and RS phase of multi-species spherical spin glasses
• Long random matrices and tensor unfolding
• Parisi formula for balanced Potts spin glass
• Near-optimal shattering in the Ising pure p-spin and rarity of solutions returned by stable algorithms
• The Kikuchi hierarchy and tensor PCA
• Shattering in pure spherical spin glasses
• Existence of full replica symmetry breaking for the Sherrington-Kirkpatrick model at low temperature
• Stochastic gradient descent in high dimensions for multi-spiked tensor PCA
• Hypergraph counting and fluctuations of mixed p-spin Glass models at high temperature
• Algorithmic thresholds for tensor PCA

This page was built for publication: Phase transition in the spiked random tensor with Rademacher prior

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2328058)