Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

SIAM Journal on Scientific Computing

Quantum Reduced-Order Modeling via Variational Quantum SVD and Linear-Layer-Enhanced Quantum LSTM

Yutong Lou, Dunhui Xiao

Published Oct 28, 2026• 1 citation
Publisher

Abstract

Abstract. This paper proposes a hybrid quantum–classical reduced-order modeling (ROM) framework for unsteady fluid flows. It is based on variational quantum singular value decomposition (VQSVD) and a linear-layer enhanced quantum long short-term memory network (L-QLSTM). The proposed architecture first applies VQSVD to compress high-dimensional flow fields into a low-dimensional latent representation parameterized by reduced coefficients. The temporal fluid dynamics evolution of these coefficients is then learnt using L-QLSTM, which maps past coefficients to the next-step coefficients through variational quantum circuits with entangling layers and linear projections. The convergence of VQSVD under finite-shot stochastic gradients induced by parameter-shift measurements is analyzed. Furthermore, Lipschitz stability for L-QLSTM is established, and a multistep inference error bound accounting for VQSVD noise and model approximation is obtained. Finally, the predicted coefficients are lifted back to the full state space using the VQSVD modes. Two Numerical experiments: flow past a cylinder and flow past two cylinders demonstrate that the proposed quantum ROM achieves faster convergence and improved long-horizon stability compared with the classical ROM. Reproducibility of computational results. This paper has been awarded the “SIAM Reproducibility Badge: Code and data available” as a recognition that the authors have followed reproducibility principles valued by SISC and the scientific computing community. Code and data that allow readers to reproduce the results in this paper are available at https://github.com/EdeIgard/QROM_VQSVD-LQLSTM and in the supplementary materials ( QROM_VQSVD-LQLSTM-main.zip [90.7MB]). [Formula: see text]

DOI
10.1137/26m1844852

Related papers

Quantum Solver Using Singular Value Decomposition for Computational Fluid Dynamics

Kalyana C. Gottiparthi, Chao Lu, Muralikrishnan Gopalakrishnan Meena +2 more

Numerical solutions for fluid flow problems are challenging and have been focus of Computational Fluid Dynamics (CFD) research for past several decades. The advent of quantum computing promises exponential speedup in comparison to existing classical methods and alleviate computational constraints posed by CFD problems. Although solutions for most problems of interest in fluid dynamics using quantum computing are distant, recent advances in algorithms, software and hardware provide a path towards realizing this goal. Quantum linear solver algorithms (QLSA) such as Harrow–Hassidim–Lloyd (HHL) and Variational Quantum Linear Solver (VQLS) have been successfully implemented to solve for canonical problems such as Hele-Shaw flow. However, these algorithms still suffer to scale and address problems with ill-conditioned Jacobians. In the current paper, we alleviate these restrictions with a new quantum solver based on Singular Value Decomposition (SVD) and simulate flow past a 2D cylinder. The fidelity of the SVD based quantum solver in predicting the flow past 2D cylinder is computed along with an assessment of errors. Classical and quantum solutions for the flow are compared for different resolutions. Finally, we discuss variation in the solutions based on number of shots used.

AIAA SciTech Forum•Jan 8, 2026•0 citations

Evaluating Hybrid Quantum-Classical Models for Reduced-Order Brain Deformation Dynamics

Tao Liu, Ge He, 梁冬雨 +1 more

We evaluate hybrid quantum-classical machine learning for the reduced-order prediction of spatiotemporal brain deformation fields. To mitigate the computational intractability of high-dimensional displacement fields, we employ Proper Orthogonal Decomposition (POD) to project the data into a compact latent space. Within this framework, we formulate two distinct learning objectives: static temporal-to-latent regression and autoregressive latent state forecasting. We systematically benchmark compact classical baselines against both minimal and enhanced hybrid quantum architectures. Our results demonstrate that classical networks provide the strongest baselines in the present setting. For static regression, a classical POD-MLP outperforms all evaluated quantum variants, although an enhanced Variational Quantum Circuit (VQC) substantially improves upon a minimal VQC baseline. For temporal forecasting, a classical POD-LSTM delivers superior predictive accuracy and statistical robustness compared to an enhanced Quantum LSTM (QLSTM) across varying history windows and random initializations. Overall, this study establishes reduced-order physical field learning as a rigorous testbed for near-term QML, highlighting that while hybrid enhancements successfully recover expressivity in weak quantum circuits, classical architectures retain a definitive advantage in both fidelity and stability.

arXiv:2610.00554•Sep 30, 2026•0 citations

QLSTM: A Quantum LSTM for Real-Time Mobile RFID Tag Tracking with Velocity Estimation

Bernard Amoah, Jian Zhang, Shiwen Mao +2 more

Accurate real-time tracking of mobile RFID tags with velocity estimation remains challenging due to phase noise, multipath interference, Doppler effects, and strict latency constraints. While classical Long Short-Term Memory (LSTM) models can capture temporal correlations in RFID measurements, they often struggle with nonlinear phase dynamics and phase-velocity coupling inherent to RF propagation. This paper presents a Quantum-Enhanced LSTM (QLSTM), a hybrid quantum-classical architecture for phase-based RFID tracking with joint position and velocity estimation. The proposed approach integrates multi-reader RFID phase measurements with variational quantum circuits employing data re-uploading and structured entanglement to extract compact temporal representations, followed by a lightweight classical decoder for real-time inference. Experiments on both simulated motion trajectories and real RFID data collected using commercial readers and USRP-based SDR platforms demonstrate that QLSTM achieves up to 23% lower position error and 31% improved velocity estimation accuracy compared to classical LSTM baselines, while maintaining inference latency below 20 ms. These results indicate that quantum-enhanced temporal modeling can improve RFID tracking performance in regimes dominated by phase periodicity, Doppler-induced dynamics, and nonlinear motion.

IEEE International Conference on RFID•Jun 1, 2026•0 citations

Learning Temporal Patterns in Financial Time Series: A Comparative Study of Quantum LSTM and Quantum Reservoir Computing

Danyal Maheshwari, Gerhard Hellstern, Martin Zaefferer +2 more

This study explores quantum and classical hybrid architectures for financial time-series fore casting, focusing on Quantum Long Short-Term Memory (QLSTM) networks and Quantum Reservoir Computing (QRC), using univariate and multivariate lag structures on real financial data. We assess how lag embeddings affect predictive accuracy and robustness. Data are en coded into quantum states via amplitude encoding, enabling efficient representation of normalized lagged observations under realistic qubit constraints. The recurrent dynamics of QLSTM and the reservoir of QRC are implemented as parameterized quantum circuits, while classical optimizers train the readout and, where applicable, variational circuit parameters. We benchmark quantum models against classical LSTM and reservoir computing using common error like metrics. Our results show that, with suitable lag selection and amplitude encoding, quantum-enhanced archi tectures match classical baselines in univariate settings and can modestly outperform them in multivariate regimes with correlated inputs, where expressive encodings are most beneficial.

quant-pharXiv:2605.02656•May 5, 2026•1 citation

Variational Quantum Singular Value Decomposition

Xin Wang, Zhixin Song, Youle Wang

Singular value decomposition is central to many problems in engineering and scientific fields. Several quantum algorithms have been proposed to determine the singular values and their associated singular vectors of a given matrix. Although these algorithms are promising, the required quantum subroutines and resources are too costly on near-term quantum devices. In this work, we propose a variational quantum algorithm for singular value decomposition (VQSVD). By exploiting the variational principles for singular values and the Ky Fan Theorem, we design a novel loss function such that two quantum neural networks (or parameterized quantum circuits) could be trained to learn the singular vectors and output the corresponding singular values. Furthermore, we conduct numerical simulations of VQSVD for random matrices as well as its applications in image compression of handwritten digits. Finally, we discuss the applications of our algorithm in recommendation systems and polar decomposition. Our work explores new avenues for quantum information processing beyond the conventional protocols that only works for Hermitian data, and reveals the capability of matrix decomposition on near-term quantum devices.

arXiv:2006.02336•Quantum•Jun 29, 2021•84 citations

Improved quantum long short-term memory with successive variational mode decomposition for solar irradiance prediction

Snigdha Sarangi, P.K. Dash, Ranjeeta Bisoi

Short-term solar irradiance prediction is critical for efficient power scheduling, the design of solar-integrated energy management systems, and maintaining grid stability. However, the inherent variability of solar energy introduces fluctuations in power output, and posing challenges for reliable grid operation. Leveraging recent advancements in quantum computing, the current work proposes a hybrid quantum-driven neural network for solar irradiance prediction, combining successive variational mode decomposition (SVMD) with an improved quantum long short-term memory (IQLSTM) network. To enhance circuit expressibility, a robust variational quantum circuit (RVQC) incorporating controlled-X (CRX) gates is embedded within the quantum long short-term memory (QLSTM) architecture. In the RVQC, a circular topology with two-way connections is employed to optimize qubit interconnections and strengthen entanglement. The proposed model is evaluated using two solar irradiance datasets of Hyderabad (2019) and Bengaluru (2019, 2018), obtained from the National Renewable Energy Laboratory (NREL). To assess the effectiveness of the proposed circuit, evaluations are carried out under varying weather conditions and across multiple forecasting horizons, and results are compared against classical LSTM and other quantum-based networks like quantum gate recurrent units (QGRUs), quantum support vector machine (QSVM) and QLSTM. Furthermore, the model is evaluated under noisy quantum environments to approximate realistic Noisy Intermediate-Scale Quantum (NISQ) hardware conditions, enabling the assessment of its robustness under practical quantum computing constraints. Experimental results demonstrate that the proposed model achieves superior predictive accuracy and robustness compared with existing classical and quantum-based approaches.

Scientific Reports•May 19, 2026•0 citations
Abstract. This paper proposes a hybrid quantum–classical reduced-order modeling (ROM) framework for unsteady fluid flows. It is based on variational quantum singular value decomposition (VQSVD) and a linear-layer enhanced quantum long short-term memory network (L-QLSTM). The proposed architecture first applies VQSVD to compress high-dimensional flow fields into a low-dimensional latent representation parameterized by reduced coefficients. The temporal fluid dynamics evolution of these coefficients is then learnt using L-QLSTM, which maps past coefficients to the next-step coefficients through variational quantum circuits with entangling layers and linear projections. The convergence of VQSVD under finite-shot stochastic gradients induced by parameter-shift measurements is analyzed. Furthermore, Lipschitz stability for L-QLSTM is established, and a multistep inference error bound accounting for VQSVD noise and model approximation is obtained. Finally, the predicted coefficients are lifted back to the full state space using the VQSVD modes. Two Numerical experiments: flow past a cylinder and flow past two cylinders demonstrate that the proposed quantum ROM achieves faster convergence and improved long-horizon stability compared with the classical ROM. Reproducibility of computational results. This paper has been awarded the “SIAM Reproducibility Badge: Code and data available” as a recognition that the authors have followed reproducibility principles valued by SISC and the scientific computing community. Code and data that allow readers to reproduce the results in this paper are available at https://github.com/EdeIgard/QROM_VQSVD-LQLSTM and in the supplementary materials ( QROM_VQSVD-LQLSTM-main.zip [90.7MB]). [Formula: see text]
Numerical solutions for fluid flow problems are challenging and have been focus of Computational Fluid Dynamics (CFD) research for past several decades. The advent of quantum computing promises exponential speedup in comparison to existing classical methods and alleviate computational constraints posed by CFD problems. Although solutions for most problems of interest in fluid dynamics using quantum computing are distant, recent advances in algorithms, software and hardware provide a path towards realizing this goal. Quantum linear solver algorithms (QLSA) such as Harrow–Hassidim–Lloyd (HHL) and Variational Quantum Linear Solver (VQLS) have been successfully implemented to solve for canonical problems such as Hele-Shaw flow. However, these algorithms still suffer to scale and address problems with ill-conditioned Jacobians. In the current paper, we alleviate these restrictions with a new quantum solver based on Singular Value Decomposition (SVD) and simulate flow past a 2D cylinder. The fidelity of the SVD based quantum solver in predicting the flow past 2D cylinder is computed along with an assessment of errors. Classical and quantum solutions for the flow are compared for different resolutions. Finally, we discuss variation in the solutions based on number of shots used.
We evaluate hybrid quantum-classical machine learning for the reduced-order prediction of spatiotemporal brain deformation fields. To mitigate the computational intractability of high-dimensional displacement fields, we employ Proper Orthogonal Decomposition (POD) to project the data into a compact latent space. Within this framework, we formulate two distinct learning objectives: static temporal-to-latent regression and autoregressive latent state forecasting. We systematically benchmark compact classical baselines against both minimal and enhanced hybrid quantum architectures. Our results demonstrate that classical networks provide the strongest baselines in the present setting. For static regression, a classical POD-MLP outperforms all evaluated quantum variants, although an enhanced Variational Quantum Circuit (VQC) substantially improves upon a minimal VQC baseline. For temporal forecasting, a classical POD-LSTM delivers superior predictive accuracy and statistical robustness compared to an enhanced Quantum LSTM (QLSTM) across varying history windows and random initializations. Overall, this study establishes reduced-order physical field learning as a rigorous testbed for near-term QML, highlighting that while hybrid enhancements successfully recover expressivity in weak quantum circuits, classical architectures retain a definitive advantage in both fidelity and stability.
Accurate real-time tracking of mobile RFID tags with velocity estimation remains challenging due to phase noise, multipath interference, Doppler effects, and strict latency constraints. While classical Long Short-Term Memory (LSTM) models can capture temporal correlations in RFID measurements, they often struggle with nonlinear phase dynamics and phase-velocity coupling inherent to RF propagation. This paper presents a Quantum-Enhanced LSTM (QLSTM), a hybrid quantum-classical architecture for phase-based RFID tracking with joint position and velocity estimation. The proposed approach integrates multi-reader RFID phase measurements with variational quantum circuits employing data re-uploading and structured entanglement to extract compact temporal representations, followed by a lightweight classical decoder for real-time inference. Experiments on both simulated motion trajectories and real RFID data collected using commercial readers and USRP-based SDR platforms demonstrate that QLSTM achieves up to 23% lower position error and 31% improved velocity estimation accuracy compared to classical LSTM baselines, while maintaining inference latency below 20 ms. These results indicate that quantum-enhanced temporal modeling can improve RFID tracking performance in regimes dominated by phase periodicity, Doppler-induced dynamics, and nonlinear motion.
This study explores quantum and classical hybrid architectures for financial time-series fore casting, focusing on Quantum Long Short-Term Memory (QLSTM) networks and Quantum Reservoir Computing (QRC), using univariate and multivariate lag structures on real financial data. We assess how lag embeddings affect predictive accuracy and robustness. Data are en coded into quantum states via amplitude encoding, enabling efficient representation of normalized lagged observations under realistic qubit constraints. The recurrent dynamics of QLSTM and the reservoir of QRC are implemented as parameterized quantum circuits, while classical optimizers train the readout and, where applicable, variational circuit parameters. We benchmark quantum models against classical LSTM and reservoir computing using common error like metrics. Our results show that, with suitable lag selection and amplitude encoding, quantum-enhanced archi tectures match classical baselines in univariate settings and can modestly outperform them in multivariate regimes with correlated inputs, where expressive encodings are most beneficial.
Singular value decomposition is central to many problems in engineering and scientific fields. Several quantum algorithms have been proposed to determine the singular values and their associated singular vectors of a given matrix. Although these algorithms are promising, the required quantum subroutines and resources are too costly on near-term quantum devices. In this work, we propose a variational quantum algorithm for singular value decomposition (VQSVD). By exploiting the variational principles for singular values and the Ky Fan Theorem, we design a novel loss function such that two quantum neural networks (or parameterized quantum circuits) could be trained to learn the singular vectors and output the corresponding singular values. Furthermore, we conduct numerical simulations of VQSVD for random matrices as well as its applications in image compression of handwritten digits. Finally, we discuss the applications of our algorithm in recommendation systems and polar decomposition. Our work explores new avenues for quantum information processing beyond the conventional protocols that only works for Hermitian data, and reveals the capability of matrix decomposition on near-term quantum devices.
Short-term solar irradiance prediction is critical for efficient power scheduling, the design of solar-integrated energy management systems, and maintaining grid stability. However, the inherent variability of solar energy introduces fluctuations in power output, and posing challenges for reliable grid operation. Leveraging recent advancements in quantum computing, the current work proposes a hybrid quantum-driven neural network for solar irradiance prediction, combining successive variational mode decomposition (SVMD) with an improved quantum long short-term memory (IQLSTM) network. To enhance circuit expressibility, a robust variational quantum circuit (RVQC) incorporating controlled-X (CRX) gates is embedded within the quantum long short-term memory (QLSTM) architecture. In the RVQC, a circular topology with two-way connections is employed to optimize qubit interconnections and strengthen entanglement. The proposed model is evaluated using two solar irradiance datasets of Hyderabad (2019) and Bengaluru (2019, 2018), obtained from the National Renewable Energy Laboratory (NREL). To assess the effectiveness of the proposed circuit, evaluations are carried out under varying weather conditions and across multiple forecasting horizons, and results are compared against classical LSTM and other quantum-based networks like quantum gate recurrent units (QGRUs), quantum support vector machine (QSVM) and QLSTM. Furthermore, the model is evaluated under noisy quantum environments to approximate realistic Noisy Intermediate-Scale Quantum (NISQ) hardware conditions, enabling the assessment of its robustness under practical quantum computing constraints. Experimental results demonstrate that the proposed model achieves superior predictive accuracy and robustness compared with existing classical and quantum-based approaches.