Boltzmann machines and energy-based models
arXiv:1708.06008
Abstract
We review Boltzmann machines and energy-based models. A Boltzmann machine defines a probability distribution over binary-valued patterns. One can learn parameters of a Boltzmann machine via gradient based approaches in a way that log likelihood of data is increased. The gradient and Hessian of a Boltzmann machine admit beautiful mathematical representations, although computing them is in general intractable. This intractability motivates approximate methods, including Gibbs sampler and contrastive divergence, and tractable alternatives, namely energy-based models.
36 pages. The topics covered in this paper are presented in Part I of IJCAI-17 tutorial on energy-based machine learning. https://researcher.watson.ibm.com/researcher/view_group.php?id=7834
References in corpus (7)
- Deep Learning in Neural Networks: An Overview
- Energy-based Generative Adversarial Network
- Reinforcement Learning with Deep Energy-Based Policies
- Interpretation and Generalization of Score Matching
- Expectation-Maximization for Learning Determinantal Point Processes
- Calibrating Energy-based Generative Adversarial Networks
- Boltzmann machines for time-series