Abstract:Tensor networks are a popular and computationally efficient approach to simulate general quantum systems on classical computers and, in a broader sense, a framework for dealing with high-dimensional numerical problems. This paper presents a broad literature review of state-of-the-art applications of tensor networks and related topics across many research domains including: machine learning, mathematical optimisation, materials science, quantum chemistry and quantum circuit simulation. This review aims to clarify which classes of relevant applications have been proposed for which class of tensor networks, and how these perform compared with other classical or quantum simulation methods. We intend this review to be a high-level tour on tensor network applications which is easy to read by non-experts, focusing on key results and limitations rather than low-level technical details of tensor networks.
What problem does this paper attempt to address?
The aim of this paper is to address the problem of how to efficiently simulate quantum systems on classical computers, particularly through the use of Tensor Networks (TNs). Specifically, the paper covers the following aspects:
1. **Cross-Disciplinary Application Review**: The paper provides a review of the latest applications of TNs across multiple research fields, including machine learning, mathematical optimization, materials science, quantum chemistry, and quantum circuit simulation. The goal is to clarify which types of applications are best suited for which types of TNs and to compare the performance of these methods with other classical or quantum simulation methods.
2. **Basic Concepts of Tensor Networks**: The paper briefly introduces the basic concepts of tensors, tensor contraction, and TN representation of state vectors, so that non-specialist readers can also understand.
3. **Image Classification**: The paper discusses in detail the application of TNs in image classification, such as handwritten digit recognition (MNIST, Fashion-MNIST, and CIFAR-10 datasets), COVID-19 pneumonia detection, etc., and compares the performance of different TN structures (such as MPS, TTN, and PEPS).
4. **Optimization and Local Hamiltonians**: The paper explores the application of TNs in optimization problems, particularly for 2-local Hamiltonian problems, which are of significant importance in quantum computing. Additionally, it discusses variational quantum algorithms and the "barren plateaus" problem they encounter.
In summary, the main purpose of this paper is to provide a comprehensive and in-depth review of the applications of TNs in various fields, helping researchers better understand the advantages and limitations of TNs and providing references for their practical applications.