English

Stochastic gradient method with accelerated stochastic dynamics

Machine Learning 2016-05-04 v1 Disordered Systems and Neural Networks Statistical Mechanics Computer Vision and Pattern Recognition

Abstract

In this paper, we propose a novel technique to implement stochastic gradient methods, which are beneficial for learning from large datasets, through accelerated stochastic dynamics. A stochastic gradient method is based on mini-batch learning for reducing the computational cost when the amount of data is large. The stochasticity of the gradient can be mitigated by the injection of Gaussian noise, which yields the stochastic Langevin gradient method; this method can be used for Bayesian posterior sampling. However, the performance of the stochastic Langevin gradient method depends on the mixing rate of the stochastic dynamics. In this study, we propose violating the detailed balance condition to enhance the mixing rate. Recent studies have revealed that violating the detailed balance condition accelerates the convergence to a stationary state and reduces the correlation time between the samplings. We implement this violation of the detailed balance condition in the stochastic gradient Langevin method and test our method for a simple model to demonstrate its performance.

Keywords

Cite

@article{arxiv.1511.06036,
  title  = {Stochastic gradient method with accelerated stochastic dynamics},
  author = {Masayuki Ohzeki},
  journal= {arXiv preprint arXiv:1511.06036},
  year   = {2016}
}

Comments

12 pages, proceedings for International Meeting on High-Dimensional Data Driven Science (HD3-2015) (http://www.sparse-modeling.jp/HD3-2015/index_e.html)

R2 v1 2026-06-22T11:49:01.887Z