Learning Relevant Features of Data with Multi-scale Tensor Networks

Abstract

Inspired by coarse-graining approaches used in physics, we show how similar algorithms can be adapted for data. The resulting algorithms are based on layered tree tensor networks and scale linearly with both the dimension of the input and the training set size. Computing most of the layers with an unsupervised algorithm, then optimizing just the top layer for supervised classification of the MNIST and fashion-MNIST data sets gives very good results. We also discuss mixing a prior guess for supervised weights together with an unsupervised representation of the data, yielding a smaller number of features nevertheless able to give good performance.

12 pages, 13 figures

References in corpus (8)

• The density-matrix renormalization group in the age of matrix product states
• Kernel methods in machine learning
• Efficient quantum state tomography
• Tensor renormalization group approach to 2D classical lattice models
• The iTEBD algorithm beyond unitary evolution
• Minimally Entangled Typical Thermal State Algorithms
• Sequentially generated states for the study of two dimensional systems
• Multigrid Algorithms for Tensor Network States

Cited by in corpus (60)

• Machine learning and the physical sciences
• The ITensor Software Library for Tensor Network Calculations
• Tensor networks for complex quantum systems
• Towards Quantum Machine Learning with Tensor Networks
• Unsupervised Generative Modeling Using Matrix Product States
• Differentiable Programming Tensor Networks
• Is quantum advantage the right goal for quantum machine learning?
• Hyper-optimized tensor network contraction
• Machine Learning for Quantum Matter
• Variational Quantum Reinforcement Learning via Evolutionary Optimization
• Quantum Neural Network States: A Brief Review of Methods and Applications
• Enhancing Generative Models via Quantum Correlations
• An end-to-end trainable hybrid classical-quantum classifier
• Bayesian Deep Learning on a Quantum Computer
• Supervised Learning with Projected Entangled Pair States
• Tensor network based machine learning of non-Markovian quantum processes
• The Presence and Absence of Barren Plateaus in Tensor-network Based Machine Learning
• Matrix Product Operators for Sequence to Sequence Learning
• Drawing Phase Diagrams of Random Quantum Systems by Deep Learning the Wave Functions
• Accurate simulation for finite projected entangled pair states in two dimensions
• Density Matrix Renormalization Group with Tensor Processing Units
• Tensor networks for unsupervised machine learning
• Tree tensor network classifiers for machine learning: from quantum-inspired to quantum-assisted
• Generative machine learning with tensor networks: benchmarks on near-term quantum computers
• A programming guide for tensor networks with global $SU(2)$ symmetry
• Tensor-based algorithms for image classification
• Entanglement-guided architectures of machine learning by quantum tensor network
• Matrix Product States for Quantum Stochastic Modelling
• Modeling Sequences with Quantum States: A Look Under the Hood
• Experimental realization of a quantum image classifier via tensor-network-based machine learning
• Mutual Information Scaling for Tensor Network Machine Learning
• Quantum Compressed Sensing with Unsupervised Tensor-Network Machine Learning
• Supervised Learning with Quantum Measurements
• Tangent-Space Gradient Optimization of Tensor Network for Machine Learning
• Unsupervised Recognition of Informative Features via Tensor Network Machine Learning and Quantum Entanglement Variations
• Optimization schemes for unitary tensor-network circuit
• Fine-Grained Tensor Network Methods
• Decohering Tensor Network Quantum Machine Learning Models
• Machine learning with tree tensor networks, CP rank constraints, and tensor dropout
• Explainable Natural Language Processing with Matrix Product States
• Variational Tensor Neural Networks for Deep Learning
• Positive unlabeled learning with tensor networks
• Measures of distinguishability between stochastic processes
• The cross-sectional stock return predictions via quantum neural network and tensor network
• Privacy-preserving machine learning with tensor networks
• Stack operation of tensor networks
• Residual Matrix Product State for Machine Learning
• Interaction Decompositions for Tensor Network Regression
• Correlated states in super-moiré materials with a kernel polynomial quantics tensor cross interpolation algorithm
• The resource theory of tensor networks
• Tensor networks and efficient descriptions of classical data
• Deep tensor networks with matrix product operators
• Cons-training Tensor Networks: Embedding and Optimization Over Discrete Linear Constraints
• Image Classification by Throwing Quantum Kitchen Sinks at Tensor Networks
• Tensor-Networks-based Learning of Probabilistic Cellular Automata Dynamics
• Scaling of contraction costs for entanglement renormalization algorithms including tensor Trotterization and variational Monte Carlo
• Quantum tensor network algorithms for evaluation of spectral functions on quantum computers
• Simulating Quantum Circuits with Tree Tensor Networks using Density-Matrix Renormalization Group Algorithm
• Quantum Dimension Reduction of Hidden Markov Models
• Explaining Anomalies with Tensor Networks