Tensor networks for quantum machine learning

Hans-Martin Rieser,Frank Köster,Arne Peter Raulf

DOI: https://doi.org/10.1098/rspa.2023.0218

2023-03-21

Abstract:Once developed for quantum theory, tensor networks have been established as a successful machine learning paradigm. Now, they have been ported back to the quantum realm in the emerging field of quantum machine learning to assess problems that classical computers are unable to solve efficiently. Their nature at the interface between physics and machine learning makes tensor networks easily deployable on quantum computers. In this review article, we shed light on one of the major architectures considered to be predestined for variational quantum machine learning. In particular, we discuss how layouts like MPS, PEPS, TTNs and MERA can be mapped to a quantum computer, how they can be used for machine learning and data encoding and which implementation techniques improve their performance.

Quantum Physics,Machine Learning

What problem does this paper attempt to address?

The paper primarily explores the applications and advancements of Tensor Networks (TNs) in the field of quantum machine learning. Specifically, the paper attempts to address the following key issues:

1. **Tensor Networks as an Effective Tool for Quantum Machine Learning**: Investigating how tensor networks can be applied to quantum machine learning to solve problems that classical computers struggle to handle efficiently. Tensor networks can be easily deployed on quantum computers due to their interface nature between physics and machine learning.

2. **Evaluation of Specific Architectures**: Focusing particularly on the main architectures considered highly suitable for variational quantum machine learning, such as Matrix Product States (MPS), Projected Entangled Pair States (PEPS), Tree Tensor Networks (TTNs), and Multiscale Entanglement Renormalization Ansatz (MERA).

3. **Mapping to Quantum Computers**: Discussing how these tensor network structures can be mapped onto quantum computers and how they can be utilized for machine learning tasks and data encoding.

4. **Performance Enhancement Techniques**: Introducing implementation techniques that can improve the performance of tensor networks on quantum computers.

In summary, the core objective of this paper is to provide theoretical guidance and technical support for this emerging field by deeply analyzing tensor networks and their applications in quantum machine learning. By focusing on different layouts of tensor networks, quantum mapping methods, and related optimization techniques, the paper aims to advance the development of quantum machine learning and explore its potential in practical applications.