Barriers and Dynamical Paths in Alternating Gibbs Sampling of Restricted Boltzmann Machines

Abstract

Restricted Boltzmann Machines (RBM) are bi-layer neural networks used for the unsupervised learning of model distributions from data. The bipartite architecture of RBM naturally defines an elegant sampling procedure, called Alternating Gibbs Sampling (AGS), where the configurations of the latent-variable layer are sampled conditional to the data-variable layer, and vice versa. We study here the performance of AGS on several analytically tractable models borrowed from statistical mechanics. We show that standard AGS is not more efficient than classical Metropolis-Hastings (MH) sampling of the effective energy landscape defined on the data layer. However, RBM can identify meaningful representations of training data in their latent space. Furthermore, using these representations and combining Gibbs sampling with the MH algorithm in the latent space can enhance the sampling performance of the RBM when the hidden units encode weakly dependent features of the data. We illustrate our findings on three datasets: Bars and Stripes and MNIST, well known in machine learning, and the so-called Lattice Proteins, introduced in theoretical biology to study the sequence-to-structure mapping in proteins.

References in corpus (16)

• Self-Learning Monte Carlo Method
• Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines
• Multitasking associative networks
• Emergence of Compositional Representations in Restricted Boltzmann Machines
• Phase Diagram of Restricted Boltzmann Machines and Generalised Hopfield Networks with Arbitrary Priors
• Self-Learning Determinantal Quantum Monte Carlo Method
• Self-learning Monte Carlo with Deep Neural Networks
• Self-Learning Monte Carlo Method: Continuous-Time Algorithm
• Benchmarking inverse statistical approaches for protein structure and design with exactly solvable models
• Thermodynamics of Restricted Boltzmann Machines and related learning dynamics
• Spectral Dynamics of Learning Restricted Boltzmann Machines
• Replica Symmetry Breaking in Bipartite Spin Glasses and Neural Networks
• Self-learning Monte Carlo method with Behler-Parrinello neural networks
• Statistical Physics and Representations in Real and Artificial Neural Networks
• Mean Field Behavior of Cluster Dynamics
• Transforming Generalized Ising Model into Boltzmann Machine

Cited by in corpus (8)

• Learning a Restricted Boltzmann Machine using biased Monte Carlo sampling
• Three Learning Stages and Accuracy-Efficiency Tradeoff of Restricted Boltzmann Machines
• Generalized hetero-associative neural networks
• Parallel Learning by Multitasking Neural Networks
• Free Energy Evaluation Using Marginalized Annealed Importance Sampling
• Interpretable machine learning of amino acid patterns in proteins: a statistical ensemble approach
• Investigating the generative dynamics of energy-based neural networks
• Emergence of Nonequilibrium Latent Cycles in Unsupervised Generative Modeling