Quantum Inspired Training for Boltzmann Machines
arXiv:1507.02642
Abstract
We present an efficient classical algorithm for training deep Boltzmann machines (DBMs) that uses rejection sampling in concert with variational approximations to estimate the gradients of the training objective function. Our algorithm is inspired by a recent quantum algorithm for training DBMs. We obtain rigorous bounds on the errors in the approximate gradients; in turn, we find that choosing the instrumental distribution to minimize the alpha=2 divergence with the Gibbs state minimizes the asymptotic algorithmic complexity. Our rejection sampling approach can yield more accurate gradients than low-order contrastive divergence training and the costs incurred in finding increasingly accurate gradients can be easily parallelized. Finally our algorithm can train full Boltzmann machines and scales more favorably with the number of layers in a DBM than greedy contrastive divergence training.
Cited by in corpus (10)
- Quantum machine learning: a classical perspective
- Opportunities and challenges for quantum-assisted machine learning in near-term quantum computers
- Tomography and Generative Data Modeling via Quantum Boltzmann Training
- Quantum-assisted Helmholtz machines: A quantum-classical deep learning framework for industrial datasets in near-term devices
- Universal discriminative quantum neural networks
- Advances in quantum machine learning
- Quantum Language Processing
- Bayesian inference via rejection filtering
- Quantum Machine Learning For Classical Data
- Quantum-inspired memory-enhanced stochastic algorithms