pith. sign in

arxiv: 2605.13200 · v1 · pith:7YLZ3EL2new · submitted 2026-05-13 · 💻 cs.LG · cs.ET

A Hybrid Tucker-LSTM Tensor Network Model for SOC Prediction in Electric Vehicles

Han Wang , Ying Wang , Bing Wang This is my paper

Pith reviewed 2026-05-14 20:27 UTC · model grok-4.3

classification 💻 cs.LG cs.ET
keywords SOC predictionTucker decompositionLSTM networkselectric vehiclesbattery managementtensor networkstime series forecastinghybrid model
0
0 comments X

The pith

Tucker tensor decomposition combined with LSTM networks improves SOC prediction accuracy for electric vehicles by preserving temporal structure in compressed data.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that a hybrid model using Tucker decomposition before LSTM processing outperforms a standard LSTM baseline on real full-lifecycle EV field data. It shows this through direct metric comparisons where error rates drop substantially while maintaining the ability to forecast state of charge from inputs like voltage, current, mileage, and cell differentials. A sympathetic reader would care because conventional estimators accumulate errors over time and rely on oversimplified battery models that fail to capture real dynamics. The work therefore positions tensor-based compression as a practical way to handle high-dimensional battery time series without sacrificing predictive power.

Core claim

The central claim is that Tucker-LSTM outperforms the baseline LSTM on all metrics when applied to full-lifecycle EV field data for SOC prediction, with MSE dropping 70.5 percent from 21.07 to 6.22, MAE improving 48.7 percent from 3.37 percent to 1.73 percent, RMSE falling from 4.59 percent to 2.49 percent, and R squared rising from 0.918 to 0.976, because Tucker decomposition reduces dimensionality while maintaining the temporal structure needed for accurate forecasting.

What carries the argument

Tucker tensor decomposition applied to multi-feature battery inputs (charge status, mileage, voltage, current, cell differentials, temporal features) to compress dimensionality before LSTM processing while preserving temporal dependencies.

If this is right

  • Tensor decomposition enables direct comparison with standard LSTM by compressing high-dimensional inputs without loss of predictive fidelity for battery time series.
  • The approach opens a direction for tensor-based analytics in electric vehicle battery management systems that handle real-world dynamics better than simplified models.
  • Improved SOC estimates from the hybrid model reduce cumulative error growth over time in vehicle battery management strategies.
  • The method supports using full feature sets including cell differentials and temporal patterns that conventional estimators often simplify away.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar tensor-LSTM hybrids could extend to other high-dimensional time-series forecasting tasks in energy systems such as load prediction or remaining useful life estimation.
  • If computational overhead of the decomposition step stays low in practice, the model becomes suitable for onboard vehicle inference rather than only offline analysis.
  • The compression benefit might combine with other sequence models like transformers to further improve accuracy on noisy sensor streams from EVs.

Load-bearing premise

Tucker decomposition reduces dimensionality of the battery data while fully preserving the temporal structure and predictive information required for accurate SOC forecasting on real EV data.

What would settle it

Running the same Tucker-LSTM and baseline LSTM models on a fresh collection of full-lifecycle EV field data and finding that the hybrid model shows equal or higher error rates than the baseline would falsify the performance claim.

Figures

Figures reproduced from arXiv: 2605.13200 by Bing Wang, Han Wang, Ying Wang.

Figure 1
Figure 1. Figure 1: Comparison chart of SOC predictions between LSTM and Tucker-LSTM [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
read the original abstract

Accurate state of charge estimation is critical for the success of electric vehicle battery management strategies, but it is well known that conventional estimators suffer from two fundamental shortcomings: cumulative errors that grow over time and reliance on simplified battery models that do not reflect real world dynamics. Therefore, this paper presents a novel hybrid approach combining Tucker tensor decomposition with LSTM networks, using full - lifecycle EV field data for SOC prediction. The inputs are charge status, mileage, voltage, current, cell differentials, and temporal features. Tucker decomposition is skillfully used to reduce dimensionality while maintaining the temporal structure, hence allowing a direct, fair comparison with standard LSTM. The result is unequivocal: Tucker - LSTM outperforms the baseline on all metrics, with MSE dropping 70.5\% (from 21.07 to 6.22 ), MAE improving 48.7\% (from 3.37\% to 1.73\%), RMSE falling from 4.59\% to 2.49\%, and $R^2$ rising from 0.918 to 0.976. Since the experimental results demonstrably demonstrate that tensor decomposition compresses high-dimensional battery data very well without loss of predictive fidelity, this paper naturally opens up a new direction for tensor-based analytics in electric vehicle battery management.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a hybrid Tucker-LSTM model for state-of-charge (SOC) prediction in electric vehicles. It constructs multi-way tensors from inputs including charge status, mileage, voltage, current, cell differentials, and temporal features; applies Tucker decomposition to reduce dimensionality while preserving temporal structure; and feeds the result into an LSTM. The central claim is that this hybrid outperforms a standard LSTM baseline on full-lifecycle EV field data, with MSE reduced 70.5% (21.07 to 6.22), MAE improved 48.7% (3.37% to 1.73%), RMSE from 4.59% to 2.49%, and R² rising from 0.918 to 0.976.

Significance. If the performance delta is shown to arise specifically from Tucker decomposition faithfully compressing the data without loss of temporal predictive information, the work would supply a concrete example of tensor methods improving battery-management forecasting and could motivate further tensor-network applications to EV time-series problems.

major comments (2)
  1. [Abstract] Abstract: the headline performance gains (MSE 21.07→6.22, etc.) are reported without any description of data splitting, preprocessing, hyperparameter selection, or statistical significance testing, so it is impossible to determine whether the 70.5% MSE reduction reflects the Tucker stage or differences in experimental protocol.
  2. [Abstract] Abstract: the claim that Tucker decomposition 'reduce[s] dimensionality while maintaining the temporal structure' is not supported by any ablation, reconstruction-error metric on the time mode, or comparison of effective sequence length/feature rank before versus after decomposition; without these controls the observed improvement could equally result from implicit regularization or altered input statistics.
minor comments (2)
  1. [Abstract] Abstract: the phrase 'full - lifecycle' contains an extraneous space; 'full-lifecycle' is the conventional form.
  2. [Abstract] Abstract: the sentence 'Since the experimental results demonstrably demonstrate...' is redundant; replace with a single verb such as 'show'.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful review and constructive suggestions. We address each major comment below and have revised the manuscript to improve clarity and provide additional supporting analyses.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the headline performance gains (MSE 21.07→6.22, etc.) are reported without any description of data splitting, preprocessing, hyperparameter selection, or statistical significance testing, so it is impossible to determine whether the 70.5% MSE reduction reflects the Tucker stage or differences in experimental protocol.

    Authors: We agree that the abstract should include sufficient experimental context to allow readers to evaluate the reported gains. In the revised manuscript we have expanded the abstract to briefly state the data splitting protocol (70/30 train/test split on the full-lifecycle field data), the preprocessing steps (z-score normalization and missing-value imputation), the hyperparameter selection procedure (grid search with 5-fold cross-validation), and the statistical significance assessment (paired t-test, p < 0.01). These additions make clear that the performance delta is measured under a controlled and identical experimental protocol for both models. revision: yes

  2. Referee: [Abstract] Abstract: the claim that Tucker decomposition 'reduce[s] dimensionality while maintaining the temporal structure' is not supported by any ablation, reconstruction-error metric on the time mode, or comparison of effective sequence length/feature rank before versus after decomposition; without these controls the observed improvement could equally result from implicit regularization or altered input statistics.

    Authors: We acknowledge that the abstract claim would be more robust with explicit quantitative support. While the manuscript already shows a direct, apples-to-apples comparison of the hybrid model against a standard LSTM that receives the identical raw feature set, we accept that additional controls are warranted. We have therefore added to the revised manuscript an ablation subsection that reports (i) reconstruction error on the temporal mode after Tucker decomposition (< 4 % relative error), (ii) the effective rank of the time mode before and after decomposition, and (iii) a comparison of sequence lengths. These metrics indicate that the temporal predictive information is largely retained, thereby reducing the plausibility of alternative explanations such as implicit regularization. revision: yes

Circularity Check

0 steps flagged

No circularity; results are direct empirical comparisons on external EV data

full rationale

The manuscript describes a hybrid Tucker-LSTM architecture and reports measured performance deltas (MSE, MAE, RMSE, R²) against a plain LSTM baseline on full-lifecycle EV field data. No equations, parameter fits, or self-citations are invoked that would make the reported gains equivalent to the inputs by construction. The central claim rests on an external benchmark comparison rather than any internal reduction or renaming of known results, satisfying the self-contained criterion.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The approach rests on standard machine-learning assumptions about sequence modeling and tensor compression rather than new axioms or entities.

axioms (2)
  • domain assumption LSTM networks can capture temporal dependencies in battery sensor streams
    Invoked by the choice of LSTM as the sequence model.
  • domain assumption Tucker decomposition can compress multi-way battery data without destroying predictive temporal structure
    Central premise of the hybrid design.

pith-pipeline@v0.9.0 · 5524 in / 1089 out tokens · 42272 ms · 2026-05-14T20:27:50.208407+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

98 extracted references · 98 canonical work pages

  1. [1]

    A lithium-ion battery remaining useful life prediction model based on ceemdan data preprocessing and hssa-lstm-tcn,

    S. Qiu, B. Zhang, Y. Lv, J. Zhang, and C. Zhang, “A lithium-ion battery remaining useful life prediction model based on ceemdan data preprocessing and hssa-lstm-tcn,”World Electric Vehicle Journal, vol. 15, no. 5, 2024

    work page 2024

  2. [2]

    Mechanistically guided residual learning for battery state monitoring throughout life,

    Y. Che, Y. Zheng, J. Rhyu, J. Guo, S. Wang, R. Teodorescu, R. D. Braatz, N. L. Bot, E. Bergin, and F. Gillespie, “Mechanistically guided residual learning for battery state monitoring throughout life,”Nature Communications, 2026

    work page 2026

  3. [3]

    Analysis of state-of-charge estimation methods for li-ion batteries considering wide temperature range,

    Y. Miao, Y. Gao, X. Liu, Y. Liang, and L. Liu, “Analysis of state-of-charge estimation methods for li-ion batteries considering wide temperature range,”Energies, vol. 18, no. 5, 2025

    work page 2025

  4. [4]

    A critical review of the state estimation methods of power batteries for electric vehicles,

    Q. Zhang, H. Rong, D. Zhao, M. Pei, and X. Dong, “A critical review of the state estimation methods of power batteries for electric vehicles,”Energies, vol. 18, no. 14, 2025

    work page 2025

  5. [5]

    Explainable real-time data driven method for battery electric model reconstruction via tensor train decomposition,

    A. Ryzhov, K. Rajinovic, H. Kuhnelt, and G. M. De, “Explainable real-time data driven method for battery electric model reconstruction via tensor train decomposition,”Journal of Power Sources, no. Jan.1, p. 625, 2025

    work page 2025

  6. [6]

    Method—research on multi-scale signal processing and efficient model construction strategies in lithium-ion battery state prediction,

    Z. Gao, R. Dai, Y. Ning, and X. Guo, “Method—research on multi-scale signal processing and efficient model construction strategies in lithium-ion battery state prediction,”Journal of The Elec- trochemical Society, vol. 171, no. 12, p. 120509, 2024

    work page 2024

  7. [7]

    Research progress on state of charge estimation methods for power batteries in new energy intelligent connected vehicles,

    H. Li, H. Jia, P. Xiao, H. Jiang, and Y. Chen, “Research progress on state of charge estimation methods for power batteries in new energy intelligent connected vehicles,”Energies, vol. 18, no. 9, 2025

    work page 2025

  8. [8]

    Tensorrt powered model for ultra-fast li-ion battery capacity predic- tion on embedded devices,

    C. Zhu, J. Qian, and M. Gao, “Tensorrt powered model for ultra-fast li-ion battery capacity predic- tion on embedded devices,”Energies (19961073), vol. 17, no. 12, 2024

    work page 2024

  9. [9]

    Latent-factorization-of-tensors-incorporated battery cycle life prediction,

    M. Chen, L. Tao, J. Lou, and X. Luo, “Latent-factorization-of-tensors-incorporated battery cycle life prediction,”IEEE/CAA Journal of Automatica Sinica, vol. 12, no. 3, pp. 633–635, 2025

    work page 2025

  10. [10]

    Review of state-of-charge estimation methods for electric vehicle applications,

    M. A. Pisani Orta, D. Garc´ ıa Elvira, and H. Valderrama Blav´ ı, “Review of state-of-charge estimation methods for electric vehicle applications,”World Electric Vehicle Journal, vol. 16, no. 2, 2025

    work page 2025

  11. [11]

    Review of state-of-charge estimation methods for electric vehicle applications,

    ——, “Review of state-of-charge estimation methods for electric vehicle applications,”World Electric Vehicle Journal, vol. 16, no. 2, 2025

    work page 2025

  12. [12]

    Electric meter,

    T. A. Edison, “Electric meter,” U.S. Patent US 223,898, Jan., 1880, filed January 27, 1880, issued January 27, 1881

    work page

  13. [13]

    Mercury electrochemical coulometer as a battery state- of-charge indicator,

    R. R. Secunde and A. G. Birchenough, “Mercury electrochemical coulometer as a battery state- of-charge indicator,” National Aeronautics and Space Administration (NASA), Washington, D.C., Tech. Rep., 1970

    work page 1970

  14. [14]

    State-of-the-art of battery state-of-charge determi- nation,

    V. Pop, H. J. Bergveld, P. H. L. Nottenet al., “State-of-the-art of battery state-of-charge determi- nation,”Measurement Science and Technology, vol. 16, no. 12, pp. R93–R110, 2005

    work page 2005

  15. [15]

    A new approach to linear filtering and prediction problems,

    R. E. Kalman, “A new approach to linear filtering and prediction problems,”Journal of Basic Engineering, vol. 82, no. 1, pp. 35–45, 1960. 8

    work page 1960

  16. [16]

    Extended kalman filtering for battery management systems of lipb-based hev battery packs. Part 2. Modeling and identification,

    G. L. Plett, “Extended kalman filtering for battery management systems of lipb-based hev battery packs. Part 2. Modeling and identification,”Journal of Power Sources, vol. 134, no. 2, pp. 262–276, 2004

    work page 2004

  17. [17]

    Electrochemical-model-based estimation of li-ion battery soc: Performance and limitations of ekf method in an electromobility use case,

    S. Ali, A. I. Bratcu, P.-X. Thivel, and I. Munteanu, “Electrochemical-model-based estimation of li-ion battery soc: Performance and limitations of ekf method in an electromobility use case,” IFAC-PapersOnLine, vol. 59, no. 9, pp. 115–120, 2025

    work page 2025

  18. [18]

    An overview of data-driven battery health estima- tion technology for battery management system,

    M. Chen, G. Ma, W. Liu, N. Zeng, and X. Luo, “An overview of data-driven battery health estima- tion technology for battery management system,”Neurocomputing, vol. 532, pp. 152–169, 2023

    work page 2023

  19. [19]

    A breif review on data-driven battery health estimation methods for energy storage systems,

    M. Chen and H. Wu, “A breif review on data-driven battery health estimation methods for energy storage systems,” in2022 IEEE International Conference on Networking, Sensing and Control (ICNSC), 2022, pp. 1–6

    work page 2022

  20. [20]

    Data-driven approaches for estimation of ev battery soc and soh: A review,

    S. G. Padder, J. Ambulkar, A. Banotra, S. Modem, S. Maheshwari, K. Jayaramulu, and C. Kundu, “Data-driven approaches for estimation of ev battery soc and soh: A review,”IEEE Access, vol. 13, pp. 35 048–35 067, 2025

    work page 2025

  21. [21]

    Neural networks-incorporated latent factor analysis for high-dimensional and incomplete data,

    M. Lin, X. Lin, X. Xu, Z. Xu, and X. Luo, “Neural networks-incorporated latent factor analysis for high-dimensional and incomplete data,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 55, no. 10, pp. 7302–7314, 2025

    work page 2025

  22. [22]

    Robust low-rank latent feature analysis for spatiotemporal signal recovery,

    D. Wu, Z. Li, Z. Yu, Y. He, and X. Luo, “Robust low-rank latent feature analysis for spatiotemporal signal recovery,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 2, pp. 2829–2842, 2025

    work page 2025

  23. [23]

    A kalman-filter-incorporated latent factor analysis model for temporally dynamic sparse data,

    Y. Yuan, X. Luo, M. Shang, and Z. Wang, “A kalman-filter-incorporated latent factor analysis model for temporally dynamic sparse data,”IEEE Transactions on Cybernetics, vol. 53, no. 9, pp. 5788–5801, 2023

    work page 2023

  24. [24]

    A fast nonnegative autoencoder-based approach to latent feature analysis on high-dimensional and incomplete data,

    F. Bi, T. He, and X. Luo, “A fast nonnegative autoencoder-based approach to latent feature analysis on high-dimensional and incomplete data,”IEEE Transactions on Services Computing, vol. 17, no. 3, pp. 733–746, 2024

    work page 2024

  25. [25]

    Multimetric autoencoder for representing high- dimensional and incomplete data,

    D. Wu, C. Liang, Y. He, Y. Qiao, and X. Luo, “Multimetric autoencoder for representing high- dimensional and incomplete data,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 56, no. 3, pp. 1533–1546, 2026

    work page 2026

  26. [26]

    Graph linear convolution pooling for learning in incomplete high-dimensional data,

    F. Bi, T. He, Y.-S. Ong, and X. Luo, “Graph linear convolution pooling for learning in incomplete high-dimensional data,”IEEE Transactions on Knowledge and Data Engineering, vol. 37, no. 4, pp. 1838–1852, 2025

    work page 2025

  27. [27]

    Nonnegative latent factor analysis-incorporated and feature-weighted fuzzy doublec-means clustering for incomplete data,

    Y. Song, M. Li, Z. Zhu, G. Yang, and X. Luo, “Nonnegative latent factor analysis-incorporated and feature-weighted fuzzy doublec-means clustering for incomplete data,”IEEE Transactions on Fuzzy Systems, vol. 30, no. 10, pp. 4165–4176, 2022

    work page 2022

  28. [28]

    Proximal alternating-direction-method-of- multipliers-incorporated nonnegative latent factor analysis,

    F. Bi, X. Luo, B. Shen, H. Dong, and Z. Wang, “Proximal alternating-direction-method-of- multipliers-incorporated nonnegative latent factor analysis,”IEEE/CAA Journal of Automatica Sinica, vol. 10, no. 6, pp. 1388–1406, 2023

    work page 2023

  29. [29]

    A 3d convolution-incorporated dimension preserved decomposition model for traffic data prediction,

    M. Lin, J. Liu, H. Chen, X. Xu, X. Luo, and Z. Xu, “A 3d convolution-incorporated dimension preserved decomposition model for traffic data prediction,”IEEE Transactions on Intelligent Trans- portation Systems, vol. 26, no. 1, pp. 673–690, 2025. 9

    work page 2025

  30. [30]

    From data analysis to intelligent maintenance: a survey on visual defect detection in aero-engines,

    P. Wu, H. Li, X. Luo, L. Hu, R. Yang, and N. Zeng, “From data analysis to intelligent maintenance: a survey on visual defect detection in aero-engines,”Measurement Science and Technology, vol. 36, no. 6, p. 062001, 2025

    work page 2025

  31. [31]

    Saliency-aware dual embedded attention network for multivariate time-series forecasting in information technology operations,

    J. Li, F. Tan, C. He, Z. Wang, H. Song, P. Hu, and X. Luo, “Saliency-aware dual embedded attention network for multivariate time-series forecasting in information technology operations,” IEEE Transactions on Industrial Informatics, vol. 20, no. 3, pp. 4206–4217, 2024

    work page 2024

  32. [32]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,”Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997

    work page 1997

  33. [33]

    Some mathematical notes on three-mode factor analysis,

    L. Tucker, “Some mathematical notes on three-mode factor analysis,”Psychometrika, vol. 31, no. 3, pp. 279–311, 1966

    work page 1966

  34. [34]

    Neural nonnegative latent factorization of tensors model with acceleration and unconstraint,

    W. Li, M. Lin, X. Xu, L. Lin, Z. Xu, and X. Luo, “Neural nonnegative latent factorization of tensors model with acceleration and unconstraint,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 56, no. 1, pp. 164–178, 2026

    work page 2026

  35. [35]

    Learning accurate representation to nonstandard tensors via a mode-aware tucker network,

    H. Wu, Q. Wang, X. Luo, and Z. Wang, “Learning accurate representation to nonstandard tensors via a mode-aware tucker network,”IEEE Transactions on Knowledge and Data Engineering, vol. 37, no. 12, pp. 7272–7285, 2025

    work page 2025

  36. [36]

    A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data,

    X. Liao, H. Wu, and X. Luo, “A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data,”IEEE Transactions on Automation Science and Engineer- ing, vol. 22, pp. 19 525–19 537, 2025

    work page 2025

  37. [37]

    Neulft: A novel approach to nonlinear canonical polyadic decomposition on high-dimensional incomplete tensors,

    X. Luo, H. Wu, and Z. Li, “Neulft: A novel approach to nonlinear canonical polyadic decomposition on high-dimensional incomplete tensors,”IEEE Transactions on Knowledge and Data Engineering, vol. 35, no. 6, pp. 6148–6166, 2023

    work page 2023

  38. [38]

    Multi-aspect self-attending neural tucker factorization for spa- tiotemporal representation learning,

    Y. Hou, P. Tang, and X. Luo, “Multi-aspect self-attending neural tucker factorization for spa- tiotemporal representation learning,”IEEE/CAA Journal of Automatica Sinica, vol. 13, no. 4, pp. 986–988, 2026

    work page 2026

  39. [39]

    Multi-indicator latent factorization of tensors for spatio-temporal signal recovery,

    C. Yu, D. Wu, J. Chen, M. Zhou, and X. Luo, “Multi-indicator latent factorization of tensors for spatio-temporal signal recovery,” in2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS). IEEE, 2025, pp. 1–8

    work page 2025

  40. [40]

    Auto-encoding neural tucker factorization,

    P. Tang, X. Luo, and J. Woodcock, “Auto-encoding neural tucker factorization,”IEEE Transactions on Knowledge and Data Engineering, vol. 37, no. 10, pp. 5795–5807, 2025

    work page 2025

  41. [41]

    A fine-grained regularization scheme for nonnegative latent factor- ization of high-dimensional and incomplete tensors,

    H. Wu, Y. Qiao, and X. Luo, “A fine-grained regularization scheme for nonnegative latent factor- ization of high-dimensional and incomplete tensors,”IEEE Trans. Serv. Comput., vol. 17, no. 6, pp. 3006–3021, 2024

    work page 2024

  42. [42]

    Advancing non-negative latent factorization of tensors with diversified regularization schemes,

    H. Wu, X. Luo, and M. Zhou, “Advancing non-negative latent factorization of tensors with diversified regularization schemes,”IEEE Transactions on Services Computing, vol. 15, no. 3, pp. 1334–1344, 2022

    work page 2022

  43. [43]

    Attention-mechanism-based neural latent-factorization-of- tensors model,

    X. Xu, M. Lin, Z. Xu, and X. Luo, “Attention-mechanism-based neural latent-factorization-of- tensors model,”ACM Transactions on Knowledge Discovery from Data, vol. 19, no. 4, pp. 1–27, 2025. 10

    work page 2025

  44. [44]

    A generalized nesterov’s accelerated gradient-incorporated non-negative latent-factorization-of-tensors model for efficient representation to dynamic qos data,

    M. Chen, R. Wang, Y. Qiao, and X. Luo, “A generalized nesterov’s accelerated gradient-incorporated non-negative latent-factorization-of-tensors model for efficient representation to dynamic qos data,” IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 8, no. 3, pp. 2386–2400, 2024

    work page 2024

  45. [45]

    A momentum-incorporated latent factorization of tensors model for temporal-aware qos missing data prediction,

    Q. Wang, M. Chen, M. Shang, and X. Luo, “A momentum-incorporated latent factorization of tensors model for temporal-aware qos missing data prediction,”Neurocomputing, vol. 367, pp. 299– 307, 2019

    work page 2019

  46. [46]

    A sampling-neighborhood-regularized latent factorization of tensor for dynamic qos estimation,

    X. Xu, M. Lin, Z. Xu, and X. Luo, “A sampling-neighborhood-regularized latent factorization of tensor for dynamic qos estimation,”IEEE Transactions on Network and Service Management, vol. 23, pp. 1707–1722, 2026

    work page 2026

  47. [47]

    Temporal pattern-aware qos prediction by biased non- negative tucker factorization of tensors,

    P. Tang, T. Ruan, H. Wu, and X. Luo, “Temporal pattern-aware qos prediction by biased non- negative tucker factorization of tensors,”Neurocomputing, vol. 582, p. 127447, 2024

    work page 2024

  48. [48]

    Temporal pattern-aware qos prediction via biased non- negative latent factorization of tensors,

    X. Luo, H. Wu, H. Yuan, and M. Zhou, “Temporal pattern-aware qos prediction via biased non- negative latent factorization of tensors,”IEEE transactions on cybernetics, vol. 50, no. 5, pp. 1798– 1809, 2020

    work page 2020

  49. [49]

    Instance-frequency-weighted regularized, nonnegative and adaptive latent fac- torization of tensors for dynamic qos analysis,

    H. Wu and X. Luo, “Instance-frequency-weighted regularized, nonnegative and adaptive latent fac- torization of tensors for dynamic qos analysis,” in2021 IEEE International Conference on Web Services (ICWS). IEEE, 2021, pp. 560–568

    work page 2021

  50. [50]

    Efficient representation to dynamic qos data via generalized nesterov’s accelerated gradient-incorporated biased non-negative latent factorization of tensors,

    M. Chen and X. Luo, “Efficient representation to dynamic qos data via generalized nesterov’s accelerated gradient-incorporated biased non-negative latent factorization of tensors,” in2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, 2021, pp. 576–581

    work page 2021

  51. [51]

    Adjusting learning depth in nonnegative latent factorization of tensors for accurately modeling temporal patterns in dynamic qos data,

    X. Luo, M. Chen, H. Wu, Z. Liu, H. Yuan, and M. Zhou, “Adjusting learning depth in nonnegative latent factorization of tensors for accurately modeling temporal patterns in dynamic qos data,” IEEE Transactions on Automation Science and Engineering, vol. 18, no. 4, pp. 2142–2155, 2021

    work page 2021

  52. [52]

    An adaptively bias-extended non-negative latent factorization of tensors model for accurately representing the dynamic qos data,

    X. Xu, M. Lin, X. Luo, and Z. Xu, “An adaptively bias-extended non-negative latent factorization of tensors model for accurately representing the dynamic qos data,”IEEE Transactions on Services Computing, vol. 18, no. 2, pp. 603–617, 2025

    work page 2025

  53. [53]

    A convolution bias-incorporated nonnegative latent factorization of tensors model for accurate representation learning to dynamic directed graphs,

    Q. Wang, H. Wu, and X. Luo, “A convolution bias-incorporated nonnegative latent factorization of tensors model for accurate representation learning to dynamic directed graphs,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 55, no. 12, pp. 8902–8914, 2025

    work page 2025

  54. [54]

    A proximal-admm-incorporated nonnegative latent- factorization-of-tensors model for representing dynamic cryptocurrency transaction network,

    X. Liao, H. Wu, T. He, and X. Luo, “A proximal-admm-incorporated nonnegative latent- factorization-of-tensors model for representing dynamic cryptocurrency transaction network,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 55, no. 11, pp. 8387–8401, 2025

    work page 2025

  55. [55]

    A pid-incorporated latent factorization of tensors approach to dynamically weighted directed network analysis,

    H. Wu, X. Luo, M. Zhou, M. J. Rawa, K. Sedraoui, and A. Albeshri, “A pid-incorporated latent factorization of tensors approach to dynamically weighted directed network analysis,”IEEE/CAA Journal of Automatica Sinica, vol. 9, no. 3, pp. 533–546, 2022

    work page 2022

  56. [56]

    Neural latent factorization of tensors for dynamically weighted directed networks analysis,

    H. Wu, X. Luo, and M. Zhou, “Neural latent factorization of tensors for dynamically weighted directed networks analysis,” in2021 IEEE International Conference on Systems, Man, and Cyber- netics (SMC). IEEE, 2021, pp. 3061–3066. 11

    work page 2021

  57. [57]

    Discovering hidden pattern in large-scale dynamically weighted directed network via latent factorization of tensors,

    ——, “Discovering hidden pattern in large-scale dynamically weighted directed network via latent factorization of tensors,” in2021 IEEE 17th International Conference on Automation Science and Engineering (CASE), 2021, pp. 1533–1538

    work page 2021

  58. [58]

    Proportional-integral-derivative-incorporated latent factorization of tensors for large-scale dynamic network analysis,

    H. Wu, Y. Xia, and X. Luo, “Proportional-integral-derivative-incorporated latent factorization of tensors for large-scale dynamic network analysis,” in2021 China Automation Congress (CAC). IEEE, 2021, pp. 2980–2984

    work page 2021

  59. [59]

    Advanced high-order graph convolutional networks with assorted time-frequency transforms,

    L. Wang, Y. Yuan, and X. Luo, “Advanced high-order graph convolutional networks with assorted time-frequency transforms,”IEEE/CAA Journal of Automatica Sinica, vol. 13, no. 2, pp. 394–408, 2026

    work page 2026

  60. [60]

    Graph tensor convolutional network,

    ——, “Graph tensor convolutional network,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 56, no. 5, pp. 3008–3024, 2026

    work page 2026

  61. [61]

    An adaptive neighborhood-resonated graph convolution network for undirected weighted graph representation,

    J. Chen, Y. Yuan, X. Luo, and X. Gao, “An adaptive neighborhood-resonated graph convolution network for undirected weighted graph representation,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 11, pp. 19 939–19 950, 2025

    work page 2025

  62. [62]

    Enhancing graph convolutional networks with an efficient k-hop neighborhood approach,

    J. Chen, X. Luo, Y. Yuan, and Z. Wang, “Enhancing graph convolutional networks with an efficient k-hop neighborhood approach,”Information Fusion, vol. 124, p. 103297, 2025

    work page 2025

  63. [63]

    Fourth-order dimension preserved tensor completion with temporal constraint for missing traffic data imputation,

    H. Chen, M. Lin, L. Zhao, Z. Xu, and X. Luo, “Fourth-order dimension preserved tensor completion with temporal constraint for missing traffic data imputation,”IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 5, pp. 6734–6748, 2025

    work page 2025

  64. [64]

    Low-rank high-order tensor com- pletion with applications in visual data,

    W. Qin, H. Wang, F. Zhang, J. Wang, X. Luo, and T. Huang, “Low-rank high-order tensor com- pletion with applications in visual data,”IEEE Transactions on Image Processing, vol. 31, pp. 2433–2448, 2022

    work page 2022

  65. [65]

    Adaptive pid-incorporated nonnegative latent factor analysis,

    J. Li, Y. Yuan, T. He, and X. Luo, “Adaptive pid-incorporated nonnegative latent factor analysis,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, pp. 1–14, 2026

    work page 2026

  66. [66]

    Genetic algorithm-based two-step optimization for precise latent factor analysis,

    C. Lyu, J. Cheng, X. Luo, and Y. Shi, “Genetic algorithm-based two-step optimization for precise latent factor analysis,”IEEE Transactions on Neural Networks and Learning Systems, vol. 37, no. 5, pp. 2294–2306, 2026

    work page 2026

  67. [67]

    Dynamic stochastic reorientation particle swarm optimization for adaptive latent factor analysis in high-dimensional sparse matrices,

    C. Lyu, Z. Ma, X. Luo, and Y. Shi, “Dynamic stochastic reorientation particle swarm optimization for adaptive latent factor analysis in high-dimensional sparse matrices,”IEEE Transactions on Knowledge and Data Engineering, vol. 38, no. 1, pp. 222–234, 2026

    work page 2026

  68. [68]

    A proportional integral controller-enhanced non-negative latent factor analysis model,

    Y. Yuan, S. Lu, and X. Luo, “A proportional integral controller-enhanced non-negative latent factor analysis model,”IEEE/CAA Journal of Automatica Sinica, vol. 12, no. 6, pp. 1246–1259, 2025

    work page 2025

  69. [69]

    Pseudo gradient-adjusted particle swarm optimization for accurate adaptive latent factor analysis,

    X. Luo, J. Chen, Y. Yuan, and Z. Wang, “Pseudo gradient-adjusted particle swarm optimization for accurate adaptive latent factor analysis,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 54, no. 4, pp. 2213–2226, 2024

    work page 2024

  70. [70]

    Position-transitional particle swarm optimization- incorporated latent factor analysis,

    X. Luo, Y. Yuan, S. Chen, N. Zeng, and Z. Wang, “Position-transitional particle swarm optimization- incorporated latent factor analysis,”IEEE Transactions on Knowledge and Data Engineering, vol. 34, no. 8, pp. 3958–3970, 2022. 12

    work page 2022

  71. [71]

    Federated latent factorization of tensors for privacy- preserving representation learning to large-scale dynamic weighted directed network,

    D. Wu, S. Zhong, Y. He, X. Luo, and X. Gao, “Federated latent factorization of tensors for privacy- preserving representation learning to large-scale dynamic weighted directed network,”IEEE Trans- actions on Dependable and Secure Computing, pp. 1–16, 2026

    work page 2026

  72. [72]

    A robust approach to electricity theft detection via tensor representation-driven contrastive distillation,

    W. Qin, Y. Ding, and X. Luo, “A robust approach to electricity theft detection via tensor representation-driven contrastive distillation,”IEEE Transactions on Industrial Informatics, pp. 1–10, 2026

    work page 2026

  73. [73]

    Tracehg: An unsupervised dual-view framework for microservice anomaly detection,

    N. Han, S. Lu, Z. Lin, B. Li, N. Wang, and X. Luo, “Tracehg: An unsupervised dual-view framework for microservice anomaly detection,”IEEE Transactions on Services Computing, vol. 19, no. 2, pp. 1633–1646, 2026

    work page 2026

  74. [74]

    Federated deep latent factor model for privacy- preserving recommendation,

    J. Gao, D. Wu, J. Chen, M. Zhou, and X. Luo, “Federated deep latent factor model for privacy- preserving recommendation,” in2025 IEEE International Conference on Systems, Man, and Cy- bernetics (SMC). IEEE, 2025, pp. 1689–1694

    work page 2025

  75. [75]

    Adversarial domain adaptation for accurately predict- ing the associations between herbal compounds and target proteins,

    Y. Qiao, L. Hu, J. Zhang, P. Hu, and X. Luo, “Adversarial domain adaptation for accurately predict- ing the associations between herbal compounds and target proteins,” in2025 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, 2025, pp. 3072–3077

    work page 2025

  76. [76]

    Advancing healthcare with large language models: Techniques and application,

    Z. Hu, Z. Peng, Z. Bi, Q. Shen, Z. Liu, J. Lou, and X. Luo, “Advancing healthcare with large language models: Techniques and application,”IEEE/CAA Journal of Automatica Sinica, vol. 12, no. 12, pp. 2371–2398, 2025

    work page 2025

  77. [77]

    Biased block term tensor decomposition for temporal pattern-aware qos prediction,

    Q. Wang, X. Liao, and H. Wu, “Biased block term tensor decomposition for temporal pattern-aware qos prediction,”International Journal of Pattern Recognition and Artificial Intelligence, vol. 39, no. 02, p. 2550001, 2025

    work page 2025

  78. [78]

    An adaptive temporal-dependent tensor low-rank representation model for dynamic communication network embedding,

    X. Liao, Q. Hu, and P. Tang, “An adaptive temporal-dependent tensor low-rank representation model for dynamic communication network embedding,” in2024 International Conference on Net- working, Sensing and Control (ICNSC). IEEE, 2024, pp. 1–6

    work page 2024

  79. [79]

    H. Wu, X. Wu, and X. Luo,Dynamic Network Representation Based on Latent Factorization of Tensors. Springer, 2023

    work page 2023

  80. [80]

    Sgd-dyg: Self-reliant global dependency apprehending on dynamic graphs,

    M. Han, L. Wang, Y. Yuan, and X. Luo, “Sgd-dyg: Self-reliant global dependency apprehending on dynamic graphs,” inProceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 2. Association for Computing Machinery, 2025, pp. 802–813

    work page 2025

Showing first 80 references.