A Deep Learning Framework for Sequence Mining with Bidirectional LSTM and Multi-Scale Attention

Honghui Xin; Minggu Wei; Tao Yang; Yaokun Ren; Yu Cheng; Yujia Lou

REVIEW 4 major objections 5 minor 4 cited by

A Deep Learning Framework for Sequence Mining with Bidirectional LSTM and Multi-Scale Attention

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Pairing a bidirectional LSTM with multi-scale attention reaches 94.27% accuracy on a 20-class gesture dataset, beating four recent sequence classification baselines.

desk verdict The architecture is a plausible baseline, but the paper's only empirical support rests on a dataset that appears not to exist in the cited archive. read the letter →

arxiv 2504.15223 v1 pith:DR34QVQT submitted 2025-04-21 cs.LG

Tao Yang , Yu Cheng , Yaokun Ren , Yujia Lou , Minggu Wei , Honghui Xin This is my paper · ORCID

classification cs.LG

keywords BidirectionalLSTMmulti-scaleattentionsequencepatternminingtimeseriesclassificationmultivariategesturerecognitiondeeplearning

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that combining a bidirectional LSTM with attention heads that operate at several different window sizes lets a sequence classifier capture both long-range context and local key patterns at the same time. On a publicly available 20-class multivariate gesture dataset, the hybrid model reports 94.27% accuracy, along with higher precision and recall than four recent sequence-modeling baselines. The authors take this as evidence that bidirectional temporal modeling plus multi-scale weighted attention is a strong, transferable recipe for complex sequence mining. If the result holds, it gives practitioners a relatively simple architecture to try before reaching for heavier Transformer backbones.

What carries the argument

The model's engine is a two-stage pipeline. First, a BiLSTM turns the input sequence $x_1,\dots,x_T$ into hidden states $h_t=[\overrightarrow{h}_t;\overleftarrow{h}_t]$ that summarize forward and backward context. Second, $S$ attention heads, each with its own window width $w_s$, compute normalized weights $\alpha_t^{(s)}=\exp(e_t^{(s)})/\sum_{k=t-w_s}^{t+w_s}\exp(e_k^{(s)})$ with $e_t^{(s)}=\tanh(W_s h_t + b_s)$, and form per-scale context vectors $c^{(s)}=\sum_t \alpha_t^{(s)} h_t$. The concatenated vector $c=[c^{(1)};\dots;c^{(S)}]$ is passed to a fully connected layer for classification. The different window widths are what let one model attend to local and global structure simultaneously.

What would settle it

Check the cited multivariate time series archive for a dataset named Learning Gesture; if no such dataset exists with 20 classes, 5 channels, and sequence lengths of 100 to 200, the accuracy comparison has no valid test bed. Even with the dataset confirmed, re-running all four baselines on the same split with a documented hyperparameter search would settle whether the 94.27% result comes from the architecture or from the experimental setup.

Watch

Extended reading notes

Core claim

The central discovery claimed is an architectural combination rather than a new mathematical result: feeding the concatenated forward and backward hidden states of a BiLSTM into several attention heads, each restricted to a different window width, and then concatenating the resulting context vectors produces better multi-class sequence classification than the compared Transformer-style baselines on the tested multivariate gesture data. The paper also reports two empirical regularities: performance peaks near sequence length 100 and degrades for very short or very long inputs, and attention window size matters, with width 7 giving the best results while width 11 hurts accuracy.

Load-bearing premise

The paper's claims stand or fall on the assumption that the Learning Gesture dataset is the real 20-class, 5-channel archive entry described here, with sequences of length 100 to 200, and that the four baseline models were trained and tuned as fairly as the proposed model.

Editorial extensions

If this is right

On the tested data, combining BiLSTM with multi-scale attention raises accuracy to 94.27%, so similar hybrids are worth evaluating before defaulting to Transformer backbones for short to medium multivariate sequences.
Sequence length is a real factor in the model's behavior: performance peaks near 100 time steps and drops off for both shorter and longer inputs, so trimming or padding inputs to a moderate length is a cheap way to improve results.
Attention window width is a meaningful hyperparameter: widths from 3 to 7 help, while width 11 hurts, so tuning window size is part of using the method successfully.
Because the output layer is task-agnostic, the same BiLSTM-plus-multi-scale-attention stack can be dropped into other sequence classification or regression problems beyond gesture recognition.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper varies one window size at a time rather than searching combinations; a natural next test is to vary the set of windows jointly or make the window widths learnable, which could shift the reported optimal configuration.
The claimed margin over the baselines would be more convincing if the same data split, training budget, and hyperparameter search were applied to every method; a reader can test this by re-running the comparison with documented settings and standardized baselines.
Because the model is structurally lighter than Transformer variants, it may be especially practical in low-resource or real-time settings where inference cost matters, but the paper does not report runtime or parameter counts, so a latency-based comparison would be an informative extension.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper is not ready for serious consideration, because the entire empirical case is a single table built on a dataset I cannot find. The authors call it "Learning Gesture" from the UEA Multivariate Time Series Classification Archive, but that archive has no such dataset. No URL, DOI, or identifier is given. As it stands, the reported 94.27% accuracy is unfalsifiable.

What the paper does well, and it is not much: the architecture is clearly described. BiLSTM followed by S windowed attention heads and a fully connected classifier is a straightforward, sensible baseline design, and the equations for attention weighting and cross-entropy training are complete enough to reimplement. The sensitivity analysis for sequence length and window size is a reasonable thing to check, even if the execution is thin. No credit beyond that.

The soft spots are proportionately severe. First, the dataset problem is load-bearing, not a minor omission. Second, there are no error bars, no significance tests, no code, and no hyperparameter details. Table 1 reports four significant digits from what looks like a single run. Third, the baselines (Informer, TimesNet, FEDformer, TSMixer) are forecasting models; the paper does not document how they were adapted for classification or tuned fairly. Fourth, the headline configuration appears to be selected from the sensitivity analyses on the same data, so 94.27% is a fitted maximum, not an independent evaluation. Fifth, the conclusion overclaims "robustness and generalization" from one dataset. Minor: TSMixer is cited as [118] in Table 1, and some references are tangentially related.

The writing is competent and the method itself is not incoherent, but the empirical center does not hold. If the dataset turned out to be real and properly identified, I would see this as a modest incremental baseline; as submitted, it is a reproducibility failure and a citation mismatch.

My recommendation: do not send this to peer review. Return it to the authors with a request to identify the dataset traceably, release code and data, report variance and significance, document baseline adaptation, and stop selecting the best configuration on the test data. If they provide all that, a resubmission would deserve a serious referee. Until then, it is not a useful contribution to the literature.

Referee Report

4 major / 5 minor

Summary. The paper proposes a sequence classification architecture that combines a Bidirectional LSTM with a multi-scale attention mechanism. Section II gives the model formulation: a BiLSTM layer produces hidden states that are fed into S attention heads with different window sizes, and the concatenated context vectors pass through a fully connected layer for classification. The method is evaluated on a dataset referred to as 'Learning Gesture' from the UEA Multivariate Time Series Classification Archive, compared against Informer, TimesNet, FEDformer, and TSMixer in Table 1, where the proposed model is reported to reach 94.27% accuracy. The paper also includes sensitivity analyses of sequence length and attention window size in Figures 2 and 3.

Significance. If the reported results were reproducible, the contribution would be a modest empirical improvement on a single multivariate time series classification benchmark; the architecture itself is a straightforward combination of established components (BiLSTM and multi-scale attention). The paper does provide a clear mathematical description of the proposed architecture in Section II, which is a strength. However, the empirical basis is not verifiable as reported: the dataset cannot be located in the cited archive, no code or data artifact is provided, and the single comparison table reports no variance or significance tests. Because the central claim is entirely empirical, the unverifiable dataset and evaluation protocol are decisive weaknesses.

major comments (4)

[Section III-A, Table 1] The dataset 'Learning Gesture' is not listed in the UEA Multivariate Time Series Classification Archive as described in the cited reference [14]; the publicly available dataset list for that archive does not contain a dataset of this name, and the paper provides no direct URL or dataset identifier. Since Table 1 is the entirety of the empirical support for the central claim, the reported 94.27% accuracy cannot be checked or reproduced. This is a load-bearing gap in the manuscript.
[Section III-B, Table 1] No variance, number of runs, or significance tests are reported for any entry in Table 1. The statement that the proposed model's accuracy is 'significantly higher' than Informer and TimesNet is therefore unsupported statistically. For a claim of superiority over recent baselines, the paper should report means and standard deviations over multiple seeds and provide paired significance tests.
[Section III-B, Figures 2 and 3] The sensitivity analyses select the best sequence length (100) and window size (7) from performance curves that appear to be computed on the same test data used for the headline result; the paper does not state that a validation split was used for this selection. If the test set was used for tuning, the reported 94.27% is a fitted maximum rather than an independent estimate, which would invalidate the comparison with baselines that were not tuned in the same manner.
[Section III-A, Table 1] The four baselines (Informer, TimesNet, FEDformer, TSMixer) are originally designed for long-term forecasting tasks. The paper does not document how these models were adapted for classification, what hyperparameters were used, or whether they were tuned on the same data. Without this information, the comparison in Table 1 cannot be interpreted as a fair evaluation of the proposed architecture.

minor comments (5)

[Table 1] The TSMixer reference appears as '[118]' rather than '[18]' in the table caption or model column; this should be corrected.
[Section II] The equations are unnumbered, which makes it difficult to refer to specific formulas; adding equation numbers would improve clarity.
[References [15]] The citation for Informer is a survey paper; the original Informer paper (Zhou et al., AAAI 2021) should be cited instead so that readers can locate the baseline.
[Section III-A] The statement that 'the average sequence length ranges from 100 to 200 time steps' is unclear; it should specify whether sequences vary between 100 and 200 time steps and how the training and test sets were split.
[Section III-B, Figures 2 and 3] The sensitivity analyses are described only in prose; no numerical values, axis labels, or error bars are provided, so the claimed trends cannot be assessed quantitatively.

Circularity Check

1 steps flagged · score 6.0 of 10

The reported 94.27% accuracy is the test-set-tuned maximum from the window-size sensitivity analysis, making the headline 'best performance' a fitted result rather than an independent evaluation.

fitted input called prediction [Section III-B, Figure 3 sensitivity analysis (window size) and Table 1 accuracy comparison]
"When the window size is set to 7, all three metrics reach their highest values. ... Therefore, window size should be carefully tuned based on the characteristics of the data to achieve optimal mining performance and generalization ability. ... The accuracy reaches 94.27%"

The paper's sensitivity analysis (Figure 3) is conducted on the same 'Learning Gesture' dataset as the accuracy comparison, and it reports that window size 7 yields the highest accuracy, precision, and recall. The paper then recommends tuning the window size on the data 'to achieve optimal mining performance' and, in the comparison, reports the proposed model's accuracy as 94.27%, claiming it 'achieves the best performance.' Since the reported configuration is defined as the one that maximizes the same metrics on the same data, the headline accuracy is a selected maximum rather than an independent result.

full rationale

The derivation chain in Section II is not circular: the BiLSTM and multi-scale attention equations are standard feed-forward computations with no hidden dependence on the experimental outcomes. No theorem is derived from its conclusion, and no result is imported from a self-citation chain: the only self-citation ([5], by coauthor Y. Cheng) supports the generic capability of BiLSTM and is not load-bearing. The central circularity is experimental. Section III-B first reports the proposed model's 94.27% accuracy in Table 1, then, in Figure 3, performs a sensitivity analysis of the attention window size on the same dataset and states that w=7 gives the highest values of all three metrics. If Table 1's configuration is the w=7 optimum (as the paper's recommendation to tune window size 'to achieve optimal mining performance' implies), then the headline accuracy is the selected maximum of the sensitivity curve, not an independent evaluation. This is the classic test-set double-dipping pattern: the same data are used to choose the hyperparameter and to measure the claim. The comparison is also asymmetric because the baselines' hyperparameter tuning is not documented. The paper's other major weakness, that 'Learning Gesture' cannot be found in the cited UEA archive, is a reproducibility/correctness problem, not a circularity, so it does not affect the score. Overall, the architecture itself is not circular; the reported performance advantage reduces to a fitted maximum, giving a partial circularity score of 6.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim is empirical and depends on dataset identity, baseline fairness, and unreported hyperparameters. No new physical or ontological entities are introduced. The architecture combines existing components.

free parameters (4)

Attention window size w_s = 7 (reported optimal)
The sensitivity analysis in Section III-B selects w=7 as the best setting on the same dataset; the final model likely uses it, so the reported accuracy is a selected maximum.
Input sequence length T = 100 (reported optimal)
The robustness experiment in Figure 2 shows the best metrics at length 100; if this value is chosen from test data, the result is not an independent evaluation.
BiLSTM hidden size and training hyperparameters = Not reported
The paper never states hidden dimension, learning rate, epochs, batch size, or optimizer, all of which are required to reproduce the model.
Number of attention scales S and per-scale window sizes = Not reported
Section II defines S scales but the experiments do not say how many scales were used or what values of w_s were combined in the final model.

assumptions (4)

domain assumption The Learning Gesture dataset exists in the UEA archive as described.
Section III-A relies on this for every experiment; the cited archive paper [14] does not list a dataset of this name.
domain assumption The four baseline models were fairly adapted and tuned for multi-class classification.
Table 1 compares against models designed primarily for forecasting; the paper does not describe adaptation or hyperparameter tuning.
domain assumption Sequence pattern mining is represented by supervised classification.
Section II defines the task as softmax classification with cross-entropy, so the claimed mining capability is measured only by classification accuracy.
standard math Standard softmax, LSTM, and attention equations from the literature are valid.
Section II uses these equations without proof; they are accepted background.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Deep Learning Framework for Sequence Mining with Bidirectional LSTM and Multi-Scale Attention." pith.science (2026). https://pith.science/paper/DR34QVQT

@misc{pith2026250415223,
  author       = {Pith},
  title        = {Pith review of: A Deep Learning Framework for Sequence Mining with Bidirectional LSTM and Multi-Scale Attention},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DR34QVQT}},
  note         = {Machine review of arXiv:2504.15223}
}

read the original abstract

This paper addresses the challenges of mining latent patterns and modeling contextual dependencies in complex sequence data. A sequence pattern mining algorithm is proposed by integrating Bidirectional Long Short-Term Memory (BiLSTM) with a multi-scale attention mechanism. The BiLSTM captures both forward and backward dependencies in sequences, enhancing the model's ability to perceive global contextual structures. At the same time, the multi-scale attention module assigns adaptive weights to key feature regions under different window sizes. This improves the model's responsiveness to both local and global important information. Extensive experiments are conducted on a publicly available multivariate time series dataset. The proposed model is compared with several mainstream sequence modeling methods. Results show that it outperforms existing models in terms of accuracy, precision, and recall. This confirms the effectiveness and robustness of the proposed architecture in complex pattern recognition tasks. Further ablation studies and sensitivity analyses are carried out to investigate the effects of attention scale and input sequence length on model performance. These results provide empirical support for structural optimization of the model.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Artificial Intelligence-Based Multiscale Temporal Modeling for Anomaly Detection in Cloud Services
cs.LG 2025-08 reject novelty 3.0 of 10

A Transformer plus multiscale attention-weighted fusion is claimed to improve cloud anomaly detection metrics by 2-3 points, but the missing label definition and artifacts block verification.
Graph Neural Network-Based Collaborative Perception for Adaptive Scheduling in Distributed Systems
cs.LG 2025-05 reject novelty 3.0 of 10

On a private simulated scheduling benchmark, a GNN with message passing and global-local fusion reports higher task completion and lower latency than four baselines, without released code, data, or error bars.
Deep Probabilistic Modeling of User Behavior for Anomaly Detection via Mixture Density Networks
cs.LG 2025-05 reject novelty 2.0 of 10

A mixture density network with negative log-likelihood scoring is claimed to outperform neural baselines on UNSW-NB15, but the experimental support is not auditable.
Joint Graph Convolution and Sequential Modeling for Scalable Network Traffic Estimation
cs.LG 2025-05 reject novelty 2.0 of 10

The paper reports that a GCN+GRU model achieves MAE 2.01, RMSE 4.12, and R2 0.956 on Abilene network traffic, outperforming four baselines.

Reference graph

Works this paper leans on

22 extracted references · 19 canonical work pages · cited by 4 Pith papers

[14]

TheUEAmultivariatetimeseriesclassificationarchive,2018,

A. Bagnall, H. A. Dau, J. Lines, M. Flynn, J. Large, A. Bostrom, et al., “TheUEAmultivariatetimeseriesclassificationarchive,2018,” arXiv preprintarXiv:1811.00075,2018

arXiv 2018
[1]

Sequential pattern mining algorithms and their applications: a technical review,

N. Mazumdar and P. K. D. Sarma, “ Sequential pattern mining algorithms and their applications: a technical review, ” International JournalofDataScienceandAnalytics,pp.1–44,2024

work page 2024
[2]

User Intent Prediction and Response in Human- Computer Interaction via BiLSTM,

Q. Sun and S. Duan, “User Intent Prediction and Response in Human- Computer Interaction via BiLSTM,” Journal of Computer Science and SoftwareApplications,vol.5,no.3,2025

work page 2025
[3]

Pattern mining: Current challenges and opportunities,

P. Fournier-Viger, et al., “ Pattern mining: Current challenges and opportunities,”ProceedingsoftheInternationalConferenceonDatabase Systems for Advanced Applications, Cham: Springer International Publishing,2022

work page 2022
[4]

A review on big data basedparallel and distributed approaches of pattern mining,

S. Kumar and K. K. Mohbey, “A review on big data basedparallel and distributed approaches of pattern mining, ” Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 5, pp. 1639–1662,2022

work page 2022
[5]

A Deep Learning Framework Integrating CNN and BiLSTM for Financial Systemic Risk Analysis and Prediction

Y. Cheng, Z. Xu, Y. Chen, Y. Wang, Z. Lin and J. Liu, “A Deep Learning Framework Integrating CNN and BiLSTM for Financial Systemic Risk Analysis and Prediction, ” arXiv preprint arXiv:2502.06847,2025

work page Pith review arXiv 2025
[6]

Spatiotemporal data mining: a survey on challenges andopenproblems,

A. Hamdi, et al., “Spatiotemporal data mining: a survey on challenges andopenproblems,”ArtificialIntelligenceReview,pp.1–48,2022

work page 2022
[7]

Knowledge Discovery: Methods from data mining and machine learning,

X. Shu and Y. Ye, “Knowledge Discovery: Methods from data mining and machine learning,” Social Science Research, vol. 110, pp. 102817, 2023

work page 2023

Show all 22 references

[8]

Elastic Scheduling of Micro-Modules in Edge Computing Based on LSTM Prediction,

J. Zhan, “Elastic Scheduling of Micro-Modules in Edge Computing Based on LSTM Prediction, ” Journal of Computer Technology and Software,vol.4,no.2,2025

2025
[9]

Mining Multimodal Data with Sparse Decomposition and Adaptive Weighting,

X. Wang, “Mining Multimodal Data with Sparse Decomposition and Adaptive Weighting, ” Transactions on Computational and Scientific Methods,vol.5,no.1,2025

2025
[10]

Federated Learning for Cross-Domain Data Privacy: A Distributed Approach to SecureCollaboration,

Y. Zhang, J. Liu, J. Wang, L. Dai, F. Guo and G. Cai, “Federated Learning for Cross-Domain Data Privacy: A Distributed Approach to SecureCollaboration,”arXivpreprintarXiv:2504.00282,2025

2025 arXiv
[11]

Multimodal Data-Driven Factor Models for Stock Market Forecasting,

J. Liu, “ Multimodal Data-Driven Factor Models for Stock Market Forecasting,”JournalofComputerTechnologyandSoftware,vol.4,no. 2,2025

2025
[12]

A hybrid network congestion prediction method integrating association rules and LSTM for enhanced spatiotemporal forecasting,

Y. Deng, “A hybrid network congestion prediction method integrating association rules and LSTM for enhanced spatiotemporal forecasting,” Transactions on Computational and Scientific Methods, vol. 5, no. 2, 2025

2025
[13]

Audit Fraud Detection via EfficiencyNet with Separable Convolution and Self-Attention,

X. Du, “ Audit Fraud Detection via EfficiencyNet with Separable Convolution and Self-Attention, ” Transactions on Computational and ScientificMethods,vol.5,no.2,2025

2025
[15]

Time series analysis based on informer algorithms: A survey,

Q. Zhu, et al., “Time series analysis based on informer algorithms: A survey,”Symmetry,vol.15,no.4,pp.951,2023

2023
[16]

Timesnet: Temporal 2d-variation modeling for general timeseriesanalysis,

H. Wu, et al., “Timesnet: Temporal 2d-variation modeling for general timeseriesanalysis,”arXivpreprintarXiv:2210.02186,2022

2022 arXiv
[17]

Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting,

T. Zhou, et al., “ Fedformer: Frequency enhanced decomposed transformer for long-term series forecasting, ” Proceedings of the InternationalConferenceonMachineLearning,PMLR,2022

2022
[18]

Tsmixer: Lightweight mlp-mixer model for multivariate time series forecasting,

V. Ekambaram, A. Jati, N. Nguyen, P. Sinthong and J. Kalagnanam, “Tsmixer: Lightweight mlp-mixer model for multivariate time series forecasting, ” Proceedings of the 29th ACM SIGKDD Conference on KnowledgeDiscoveryandDataMining,pp.459–469,2023

2023
[19]

Feature fusion for basic behavior unit segmentation from video sequences,

X. Xue and T. C. Henderson, “Feature fusion for basic behavior unit segmentation from video sequences, ” Robotics and Autonomous Systems,vol.57,no.3,pp.239–248,2009

2009
[20]

Identifying fusion transcripts using next generation sequencing,

S. Kumar, S. K. Razzaq, A. D. Vo, M. Gautam and H. Li, “Identifying fusion transcripts using next generation sequencing, ” Wiley InterdisciplinaryReviews:RNA,vol.7,no.6,pp.811–823,2016

2016
[21]

A reviewonmethodsandapplicationsinmultimodaldeeplearning,

S. Jabeen, X. Li, M. S. Amin, O. Bourahla, S. Li and A. Jabbar, “A reviewonmethodsandapplicationsinmultimodaldeeplearning,”ACM Transactions on Multimedia Computing, Communications and Applications,vol.19,no.2s,pp.1–41,2023

2023
[22]

Self- Supervised Graph Neural Networks for Enhanced Feature Extraction in Heterogeneous Information Networks,

J. Wei, Y. Liu, X. Huang, X. Zhang, W. Liu and X. Yan, “Self- Supervised Graph Neural Networks for Enhanced Feature Extraction in Heterogeneous Information Networks, ” 2024 5th International ConferenceonMachineLearningandComputerApplication(ICMLCA), pp.272–276,2024

2024

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [14]

TheUEAmultivariatetimeseriesclassificationarchive,2018,

A. Bagnall, H. A. Dau, J. Lines, M. Flynn, J. Large, A. Bostrom, et al., “TheUEAmultivariatetimeseriesclassificationarchive,2018,” arXiv preprintarXiv:1811.00075,2018

arXiv 2018

[2] [1]

Sequential pattern mining algorithms and their applications: a technical review,

N. Mazumdar and P. K. D. Sarma, “ Sequential pattern mining algorithms and their applications: a technical review, ” International JournalofDataScienceandAnalytics,pp.1–44,2024

work page 2024

[3] [2]

User Intent Prediction and Response in Human- Computer Interaction via BiLSTM,

Q. Sun and S. Duan, “User Intent Prediction and Response in Human- Computer Interaction via BiLSTM,” Journal of Computer Science and SoftwareApplications,vol.5,no.3,2025

work page 2025

[4] [3]

Pattern mining: Current challenges and opportunities,

P. Fournier-Viger, et al., “ Pattern mining: Current challenges and opportunities,”ProceedingsoftheInternationalConferenceonDatabase Systems for Advanced Applications, Cham: Springer International Publishing,2022

work page 2022

[5] [4]

A review on big data basedparallel and distributed approaches of pattern mining,

S. Kumar and K. K. Mohbey, “A review on big data basedparallel and distributed approaches of pattern mining, ” Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 5, pp. 1639–1662,2022

work page 2022

[6] [5]

A Deep Learning Framework Integrating CNN and BiLSTM for Financial Systemic Risk Analysis and Prediction

Y. Cheng, Z. Xu, Y. Chen, Y. Wang, Z. Lin and J. Liu, “A Deep Learning Framework Integrating CNN and BiLSTM for Financial Systemic Risk Analysis and Prediction, ” arXiv preprint arXiv:2502.06847,2025

work page Pith review arXiv 2025

[7] [6]

Spatiotemporal data mining: a survey on challenges andopenproblems,

A. Hamdi, et al., “Spatiotemporal data mining: a survey on challenges andopenproblems,”ArtificialIntelligenceReview,pp.1–48,2022

work page 2022

[8] [7]

Knowledge Discovery: Methods from data mining and machine learning,

X. Shu and Y. Ye, “Knowledge Discovery: Methods from data mining and machine learning,” Social Science Research, vol. 110, pp. 102817, 2023

work page 2023