TensorNetwork for Machine Learning

Abstract

We demonstrate the use of tensor networks for image classification with the TensorNetwork open source library. We explain in detail the encoding of image data into a matrix product state form, and describe how to contract the network in a way that is parallelizable and well-suited to automatic gradients for optimization. Applying the technique to the MNIST and Fashion-MNIST datasets we find out-of-the-box performance of 98% and 88% accuracy, respectively, using the same tensor network architecture. The TensorNetwork library allows us to seamlessly move from CPU to GPU hardware, and we see a factor of more than 10 improvement in computational speed using a GPU.

9 pages, 8 figures. All code can be found at https://github.com/google/tensornetwork

References in corpus (4)

• A class of quantum many-body states that can be efficiently simulated
• Tensor renormalization group approach to 2D classical lattice models
• Renormalization algorithms for Quantum-Many Body Systems in two and higher dimensions
• Continuous Matrix Product States for Quantum Fields

Cited by in corpus (12)

• Tensor-based algorithms for image classification
• Hybrid quantum-classical classifier based on tensor network and variational quantum circuit
• Tensor Networks for Medical Image Classification
• Quantum Tensor Network in Machine Learning: An Application to Tiny Object Classification
• Stack operation of tensor networks
• Gleipnir: Toward Practical Error Analysis for Quantum Programs (Extended Version)
• Accuracy and Efficiency of Simplified Tensor Network Codes
• A Tensor Network Approach to Finite Markov Decision Processes
• Tensor-Train Networks for Learning Predictive Modeling of Multidimensional Data
• Tensor Network for Supervised Learning at Finite Temperature
• Multi-layered tensor networks for image classification
• SchrödingeRNN: Generative Modeling of Raw Audio as a Continuously Observed Quantum State