Machine learning by unitary tensor network of hierarchical tree structure (2017-10-13T00:00:00.000000Z)

TL;DR

This work trains two-dimensional hierarchical TNs to solve image recognition problems, using a training algorithm derived from the multi-scale entanglement renormalization ansatz, and introduces mathematical connections among quantum many-body physics, quantum information theory, and machine learning.

Abstract

The resemblance between the methods used in quantum-many body physics and in machine learning has drawn considerable attention. In particular, tensor networks (TNs) and deep learning architectures bear striking similarities to the extent that TNs can be used for machine learning. Previous results used one-dimensional TNs in image recognition, showing limited scalability and flexibilities. In this work, we train two-dimensional hierarchical TNs to solve image recognition problems, using a training algorithm derived from the multi-scale entanglement renormalization ansatz. This approach introduces mathematical connections among quantum many-body physics, quantum information theory, and machine learning. While keeping the TN unitary in the training phase, TN states are defined, which encode classes of images into quantum many-body states. We study the quantum features of the TN states, including quantum entanglement and fidelity. We find these quantities could be properties that characterize the image classes, as well as the machine learning tasks.

Authors

Ding Liu

5 papers

P. Wittek

2 papers

Shi-Ju Ran

1 papers

TL;DR

Abstract

Authors

References47 items

Towards quantum machine learning with tensor networks

Learning architectures based on quantum entanglement: a simple matrix product state algorithm for image recognition

Neural-Network Quantum States, String-Bond States, and Chiral Topological States

Duality of Graphical Models and Tensor Networks

Unsupervised Generative Modeling Using Matrix Product States

Machine Learning Spatial Geometry from Entanglement Features

Review of Tensor Network Contraction Approaches

Tensor Networks for Dimensionality Reduction and Large-scale Optimization: Part 2 Applications and Future Perspectives

Holography as deep learning

Exponential Machines

Experimental demonstration of efficient quantum state tomography of matrix product states.

Deep Learning and Quantum Entanglement: Fundamental Connections with Implications to Network Design

Neural Network Representation of Tensor Network and Chiral States.

Equivalence of restricted Boltzmann machines and tensor network states

Tensor Networks for Dimensionality Reduction and Large-scale Optimization: Part 1 Low-Rank Tensor Decompositions

Solving the quantum many-body problem with artificial neural networks

On the Expressive Power of Deep Learning: A Tensor Analysis

Experimental realization of a quantum support vector machine.

Tree Tensor Network State with Variable Tensor Order: An Efficient Multireference Method for Strongly Correlated Systems

Advances on tensor network theory: symmetries, fermions, entanglement, and holography

Advances on tensor network theory: symmetries, fermions, entanglement, and holography

A Practical Introduction to Tensor Networks: Matrix Product States and Projected Entangled Pair States

Deep learning and the renormalization group

ImageNet classification with deep convolutional neural networks

Quantum simulation

Natural and artificial atoms for quantum computation

Physics: Quantum computing

A New Scheme for the Tensor Representation

Matrix product states, projected entangled pair states, and variational renormalization group methods for quantum spin systems

Multiscale entanglement renormalization ansatz in two dimensions: quantum Ising model.

Algorithms for entanglement renormalization

Class of quantum many-body states that can be efficiently simulated.

A Fast Learning Algorithm for Deep Belief Nets

Entanglement renormalization.

Classical simulation of quantum many-body systems with a tree tensor network

Quantum computing

Efficient classical simulation of slightly entangled quantum computations.

Quantum information and computation

Quantum computing

Universal Quantum Simulators

Density matrix formulation for quantum renormalization groups.

Supervised Learning with Tensor Networks

Deep Learning

C´edric B´eny

Learning Multiple Layers of Features from Tiny Images

Visualizing Data using t-SNE

f¨ur Mathematik in den Naturwissenschaften Leipzig Hierarchical Singular Value Decomposition of Tensors

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names