by Dima Kochkov

Boltzmann Machines

Boltzmann Machines represent a class of Neural Networks that can be used for unsupervised learning. Inspired by ideas from physics and neuroscience these nets allow a simple, genuine learning rule. The learning is based on minimization of Kullback–Leibler divergence between learned probability distribution and the dataset.

Presentation Summary

link to pdf

Examples

A fairly long python tutorial on RBM example
Link to a simplistic c++ code

References

A great course of Neural Networks Coursera // soon to be removed
An exact mapping between deep learning and variational RG RBM-RG
On Hopfield Nets, for historic purposes

All Machine learning

by Will Wei · Deep Learning Partial Differential Equation (PDE)
by Will Wei · Adaptive Boosting
by Zeqian Li · Markov Decision Process and Reinforcement Learning
by Yubo 'Paul' Yang · Kriging
by Dima Kochkov · Generative Adversarial Networks
by Xiongjie Yu · Neural Network on a Tensor Train
by Will Wheeler · Reservoir Computing in the Time Domain
by Yubo 'Paul' Yang · Gibbs Sampling
by Matt Zhang · Predict Seizure with EEG
by Matt Zhang · Adaptive Boosting (AdaBoost)
by Benjamin Villalonga Correa · Introduction to Supervised Machine Learning
by Dima Kochkov · Boltzmann Machines

Yubo "Paul" Yang 2016-06-08 ALGORITHM
machine learning energy based models unsupervised learning