Density Operator Expectation Maximization

TL;DR

This paper introduces Density Operator Expectation Maximization (DO-EM), a novel framework for training quantum density operator-based latent variable models, demonstrating superior generative performance over classical probabilistic models.

Contribution

The paper develops DO-EM, the first EM framework for density operators, and introduces quantum models that outperform classical counterparts on generative tasks.

Findings

Quantum models outperform probabilistic models on generative tasks.

DO-EM effectively trains quantum latent variable models.

Quantum RBMs and QG-BRBMs show improved performance.

Abstract

Machine learning with density operators, the mathematical foundation of quantum mechanics, is gaining prominence with rapid advances in quantum computing. Generative models based on density operators cannot yet handle tasks that are routinely handled by probabilistic models. The progress of latent variable models, a broad and influential class of probabilistic unsupervised models, was driven by the Expectation-Maximization framework. Deriving such a framework for density operators is challenging due to the non-commutativity of operators. To tackle this challenge, an inequality arising from the monotonicity of relative entropy is demonstrated to serve as an evidence lower bound for density operators. A minorant-maximization perspective on this bound leads to Density Operator Expectation Maximization (DO-EM), a general framework for training latent variable models defined through density operators. Through an information-geometric argument, the Expectation step in DO-EM is shown to be the Petz recovery map. The DO-EM algorithm is applied to Quantum Restricted Boltzmann Machines, adapting Contrastive Divergence to approximate the Maximization step gradient. Quantum interleaved Deep Boltzmann Machines and Quantum Gaussian-Bernoulli Restricted Boltzmann Machines, new models introduced in this work, outperform their probabilistic counterparts on generative tasks when trained with similar computational resources and identical hyperparameters.