Abstract:Restricted Boltzmann Machines (RBMs) are effective tools for modeling complex systems and deriving insights from data. However, training these models with highly structured data presents significant challenges due to the slow mixing characteristics of Markov Chain Monte Carlo processes. In this study, we build upon recent theoretical advancements in RBM training, to significantly reduce the computational cost of training (in very clustered datasets), evaluating and sampling in RBMs in general. The learning process is analogous to thermodynamic continuous phase transitions observed in ferromagnetic models, where new modes in the probability measure emerge in a continuous manner. Such continuous transitions are associated with the critical slowdown effect, which adversely affects the accuracy of gradient estimates, particularly during the initial stages of training with clustered data. To mitigate this issue, we propose a pre-training phase that encodes the principal components into a low-rank RBM through a convex optimization process. This approach enables efficient static Monte Carlo sampling and accurate computation of the partition function. We exploit the continuous and smooth nature of the parameter annealing trajectory to achieve reliable and computationally efficient log-likelihood estimations, enabling online assessment during the training, and propose a novel sampling strategy named parallel trajectory tempering (PTT) which outperforms previously optimized MCMC methods. Our results show that this training strategy enables RBMs to effectively address highly structured datasets that conventional methods struggle with. We also provide evidence that our log-likelihood estimation is more accurate than traditional, more computationally intensive approaches in controlled scenarios. The PTT algorithm significantly accelerates MCMC processes compared to existing and conventional methods.

Privacy-Preserving Restricted Boltzmann Machine

Real-Time Keypoint Recognition Using Restricted Boltzmann Machine

Generative and Discriminative Infinite Restricted Boltzmann Machine Training

Method to Improve the Performance of Restricted Boltzmann Machines.

Cardinality restricted boltzmann machines

MapReduce Based Distributed Learning Algorithm for Restricted Boltzmann Machine

Fast training and sampling of Restricted Boltzmann Machines

Sparse Restricted Boltzmann Machine Based on Data Class Entropy

Privacy-Preserving Machine Learning Algorithms for Big Data Systems

A Precise Method for RBMs Training Using Phased Curricula

A Novel Restricted Boltzmann Machine Training Algorithm with Fast Gibbs Sampling Policy

Mining Invariance in Restricted Boltzmann Machine Via Information Geometry

Gaussian Cardinality Restricted Boltzmann Machines.

Subspace Restricted Boltzmann Machine

An Overview on Restricted Boltzmann Machines.

Towards the representational power of restricted Boltzmann machines

Privacy Budgeting for Growing Machine Learning Datasets

An Optimized Dimensionality Reduction Model for High-Dimensional Data Based on Restricted Boltzmann Machines

Elastic Restricted Boltzmann Machines for Cancer Data Analysis

Temperature based Restricted Boltzmann Machines

Improving the Generalization Ability of Restricted Boltzmann Machines Via Theta Pure Dependency.