Convergence of contrastive divergence algorithm in exponential family

DOI10.1214/17-AOS1649zbMath1400.68172arXiv1603.05729WikidataQ115517793 ScholiaQ115517793MaRDI QIDQ1991693

Bai Jiang, Tung-Yu Wu, Yifan Jin, Wing-Hung Wong

Publication date: 30 October 2018

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

Full work available at URL: https://arxiv.org/abs/1603.05729

zbMATH Keywords

convergence rateexponential familycontrastive divergence

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Learning and adaptive systems in artificial intelligence (68T05) Applications of Markov chains and discrete-time Markov processes on general state spaces (social mobility, learning theory, industrial processes, etc.) (60J20)

Related Items

Learning large \(Q\)-matrix by restricted Boltzmann machines

Uses Software

• ergm
• darch
• STL-10 dataset

Cites Work

• Unnamed Item
• Unnamed Item
• A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
• General state space Markov chains and MCMC algorithms
• Geometric ergodicity and hybrid Markov chains
• Using contrastive divergence to seed Monte Carlo MLE for exponential-family random graph models
• Rates of convergence of the Hastings and Metropolis algorithms
• Convergence properties of the Gibbs sampler for perturbations of Gaussians
• Geometric ergodicity and the spectral gap of non-reversible Markov chains
• Weighted sums of certain dependent random variables
• Yet Another Look at Harris’ Ergodic Theorem for Markov Chains
• Reducing the Dimensionality of Data with Neural Networks
• Stability of Markovian processes III: Foster–Lyapunov criteria for continuous-time processes
• Training Products of Experts by Minimizing Contrastive Divergence
• Consistency of Pseudolikelihood Estimation of Fully Visible Boltzmann Machines
• The Markov chain Monte Carlo revolution
• Geometric convergence and central limit theorems for multidimensional Hastings and Metropolis algorithms
• Stability of Markovian processes I: criteria for discrete-time Chains
• 10.1162/jmlr.2003.4.7-8.1235
• Justifying and Generalizing Contrastive Divergence
• On perturbed proximal gradient algorithms
• Book Reviews
• A Fast Learning Algorithm for Deep Belief Nets