Abstract:Quantum computers are next-generation devices that hold promise to perform calculations beyond the reach of classical computers. A leading method towards achieving this goal is through quantum machine learning, especially quantum generative learning. Due to the intrinsic probabilistic nature of quantum mechanics, it is reasonable to postulate that quantum generative learning models (QGLMs) may surpass their classical counterparts. As such, QGLMs are receiving growing attention from the quantum physics and computer science communities, where various QGLMs that can be efficiently implemented on near-term quantum machines with potential computational advantages are proposed. In this paper, we review the current progress of QGLMs from the perspective of machine learning. Particularly, we interpret these QGLMs, covering quantum circuit born machines, quantum generative adversarial networks, quantum Boltzmann machines, and quantum autoencoders, as the quantum extension of classical generative learning models. In this context, we explore their intrinsic relation and their fundamental differences. We further summarize the potential applications of QGLMs in both conventional machine learning tasks and quantum physics. Last, we discuss the challenges and further research directions for QGLMs.

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is to explore the potential of quantum generative learning models (QGLMs) in handling complex computational tasks that are difficult for classical computers to achieve. Specifically, the paper focuses on how to utilize the unique properties of quantum computers to design and implement efficient QGLMs, and evaluate the computational advantages of these models on quantum machines that can be realized in the near future. ### Problem Background With the rapid development of deep generative learning models (GLMs) in the past decade, they have achieved great success in fields such as computer vision, natural language processing, and drug discovery. However, as Moore's Law approaches its limit, classical GLMs are facing the challenge of excessive computational cost. To overcome this problem, researchers have begun to explore the application of GLMs to quantum computers, hoping to use the powerful theoretical and experimental capabilities of quantum computing to enhance the capabilities of the models. ### Paper Objectives The main objective of the paper is to review the current research progress of quantum generative learning models (QGLMs), especially to explain the working principles of these models from the perspective of machine learning and their relationship with classical generative learning models. Specifically, the paper covers the following main QGLMs: 1. **Quantum Circuit Born Machine (QCBM)** 2. **Quantum Generative Adversarial Network (QGAN)** 3. **Quantum Boltzmann Machine (QBM)** 4. **Quantum Autoencoder (QAE)** ### Main Contributions - **Classification and Comparison**: The paper conducts a detailed classification and comparison of the above four QGLMs, analyzing their internal relations and fundamental differences. - **Potential Applications**: It explores the potential applications of QGLMs in traditional machine - learning tasks and quantum physics. - **Challenges and Future Directions**: It discusses the challenges faced by QGLMs and future research directions. ### Summary By systematically reviewing and analyzing the existing QGLMs, the paper aims to help researchers in the fields of computer science and quantum physics better understand the current development status and future prospects of these models. This is the first comprehensive review of modern QGLMs from the perspective of quantum generative learning, filling the gap in related fields. ### Related Formulas Some key formulas involved in the discussion include: - **Energy Function** (for Boltzmann machines): \[ E_z = -\sum_{i < j} w_{ij} z_i z_j - \sum_i b_i z_i \] - **Probability of Boltzmann Distribution**: \[ p_\theta(v) = Z^{-1} \sum_h e^{-E\{v,h\}}, \quad Z = \sum_z e^{-E_z} \] - **ELBO of Variational Autoencoder**: \[ \log p_w(x) = C_{\theta,w}(x) + D_{KL}(q_\theta(z|x) \| p_w(z|x)) \] These formulas show the mathematical expression and optimization process of quantum generative learning models, providing a theoretical basis for understanding and research.

Recent Advances for Quantum Neural Networks in Generative Learning

A Quantum Machine Learning Algorithm Based on Generative Models

Power of Quantum Generative Learning

A Survey of Recent Advances in Quantum Generative Adversarial Networks

Quantum Generative Models for Data Generation

Experimental Quantum Generative Adversarial Networks for Image Generation

An efficient quantum algorithm for generative machine learning

Generative quantum machine learning via denoising diffusion probabilistic models

Quantum Deep Generative Prior with Programmable Quantum Circuits

Recent Advances of Quantum Neural Networks on the Near Term Quantum Processor

Quantum Generative Adversarial Networks: Bridging Classical and Quantum Realms

Quantum Algorithms for Feedforward Neural Networks

Variational Quantum Circuits Enhanced Generative Adversarial Network

Application of Quantum Generative Adversarial Learning in Quantum Image Processing

Quantum generative adversarial network for generating discrete distribution

Quantum Generative Diffusion Model: A Fully Quantum-Mechanical Model for Generating Quantum State Ensemble

LatentQGAN: A Hybrid QGAN with Classical Convolutional Autoencoder

Learning and Inference on Generative Adversarial Quantum Circuits

Quantum Generative Adversarial Networks for Learning and Loading Quantum Image in Noisy Environment