Randomized tensor network reservoir computing: validity and learnability phase transitions

Reservoir computing (RC) systems, traditionally based on echo state networks (ESN) or liquid state machines, have shown significant potential in dynamic temporal data modeling, such as weather forecasting and astronomical predictions. However, these frameworks are known not being applicable to quant...

Full description

Saved in:
Bibliographic Details
Main Authors: Shinji Sato, Daiki Sasaki, Chih-Chieh Chen, Kodai Shiba, Tomah Sogabe
Format: Article
Language:English
Published: IOP Publishing 2025-01-01
Series:Machine Learning: Science and Technology
Subjects:
Online Access:https://doi.org/10.1088/2632-2153/aded56
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849732533088944128
author Shinji Sato
Daiki Sasaki
Chih-Chieh Chen
Kodai Shiba
Tomah Sogabe
author_facet Shinji Sato
Daiki Sasaki
Chih-Chieh Chen
Kodai Shiba
Tomah Sogabe
author_sort Shinji Sato
collection DOAJ
description Reservoir computing (RC) systems, traditionally based on echo state networks (ESN) or liquid state machines, have shown significant potential in dynamic temporal data modeling, such as weather forecasting and astronomical predictions. However, these frameworks are known not being applicable to quantum dynamics-based RC. Tensor networks (TNs), with their efficient representation of high-dimensional quantum information and entanglement, are powerful tools for modeling correlated quantum dynamics. Introducing randomized effects into TNs, akin to randomization of recurrent connections in traditional RC, is expected to generate diverse quantum correlation patterns and provide robust, generalizable quantum-inspired dynamic models through randomization. In this work, we propose a novel randomized TN-based RC scheme, experimentally demonstrating its validity. A theoretical model selection criterion is constructed to find the optimal TNRC hyperparameters. Critical phenomena along with the phase transitions of learnability near the edge of chaos in TN RC are clearly identified. The distribution-independent universality in phase transitions observed in TN RC is captured by the newly developed learning theory and a self-consistent mean-field theory of the spin-glass type. The performance advantage of TNRC over ESN is demonstrated in several forecasting experiments. Our findings lay the groundwork for future explorations into randomized TN quantum machine learning, phase transitions in quantum RC, and the manipulation of critical phenomena in complex systems.
format Article
id doaj-art-5f68a90fd4264fabb10d4eb024265171
institution DOAJ
issn 2632-2153
language English
publishDate 2025-01-01
publisher IOP Publishing
record_format Article
series Machine Learning: Science and Technology
spelling doaj-art-5f68a90fd4264fabb10d4eb0242651712025-08-20T03:08:18ZengIOP PublishingMachine Learning: Science and Technology2632-21532025-01-016303501110.1088/2632-2153/aded56Randomized tensor network reservoir computing: validity and learnability phase transitionsShinji Sato0Daiki Sasaki1https://orcid.org/0009-0006-9601-0773Chih-Chieh Chen2https://orcid.org/0000-0003-3092-4346Kodai Shiba3Tomah Sogabe4https://orcid.org/0000-0001-9258-6130Engineering Department, The University of Electro-Communications , Chofu 182-8585, Tokyo, JapanEngineering Department, The University of Electro-Communications , Chofu 182-8585, Tokyo, JapanGrid Inc. , Minato-ku 171-0061, Tokyo, JapanGrid Inc. , Minato-ku 171-0061, Tokyo, JapanEngineering Department, The University of Electro-Communications , Chofu 182-8585, Tokyo, Japan; Grid Inc. , Minato-ku 171-0061, Tokyo, Japan; i-PERC, The University of Electro-Communications , Chofu 182-8585, Tokyo, JapanReservoir computing (RC) systems, traditionally based on echo state networks (ESN) or liquid state machines, have shown significant potential in dynamic temporal data modeling, such as weather forecasting and astronomical predictions. However, these frameworks are known not being applicable to quantum dynamics-based RC. Tensor networks (TNs), with their efficient representation of high-dimensional quantum information and entanglement, are powerful tools for modeling correlated quantum dynamics. Introducing randomized effects into TNs, akin to randomization of recurrent connections in traditional RC, is expected to generate diverse quantum correlation patterns and provide robust, generalizable quantum-inspired dynamic models through randomization. In this work, we propose a novel randomized TN-based RC scheme, experimentally demonstrating its validity. A theoretical model selection criterion is constructed to find the optimal TNRC hyperparameters. Critical phenomena along with the phase transitions of learnability near the edge of chaos in TN RC are clearly identified. The distribution-independent universality in phase transitions observed in TN RC is captured by the newly developed learning theory and a self-consistent mean-field theory of the spin-glass type. The performance advantage of TNRC over ESN is demonstrated in several forecasting experiments. Our findings lay the groundwork for future explorations into randomized TN quantum machine learning, phase transitions in quantum RC, and the manipulation of critical phenomena in complex systems.https://doi.org/10.1088/2632-2153/aded56tensor networksreservoir computingphase transitions
spellingShingle Shinji Sato
Daiki Sasaki
Chih-Chieh Chen
Kodai Shiba
Tomah Sogabe
Randomized tensor network reservoir computing: validity and learnability phase transitions
Machine Learning: Science and Technology
tensor networks
reservoir computing
phase transitions
title Randomized tensor network reservoir computing: validity and learnability phase transitions
title_full Randomized tensor network reservoir computing: validity and learnability phase transitions
title_fullStr Randomized tensor network reservoir computing: validity and learnability phase transitions
title_full_unstemmed Randomized tensor network reservoir computing: validity and learnability phase transitions
title_short Randomized tensor network reservoir computing: validity and learnability phase transitions
title_sort randomized tensor network reservoir computing validity and learnability phase transitions
topic tensor networks
reservoir computing
phase transitions
url https://doi.org/10.1088/2632-2153/aded56
work_keys_str_mv AT shinjisato randomizedtensornetworkreservoircomputingvalidityandlearnabilityphasetransitions
AT daikisasaki randomizedtensornetworkreservoircomputingvalidityandlearnabilityphasetransitions
AT chihchiehchen randomizedtensornetworkreservoircomputingvalidityandlearnabilityphasetransitions
AT kodaishiba randomizedtensornetworkreservoircomputingvalidityandlearnabilityphasetransitions
AT tomahsogabe randomizedtensornetworkreservoircomputingvalidityandlearnabilityphasetransitions