Pith. sign in

REVIEW 3 major objections 6 minor 51 references

CARE: A Cascaded Framework for Efficient and Reliable Time Series Anomaly Detection

T0 review · 3 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A lightweight pre-filter can route only uncertain samples to a heavy detector, cutting inference cost by up to 4.8x while keeping detection quality competitive.

desk verdict A credible, well-tested cascade for TSAD with a real but addressable gap: the 80% of samples scored only by the lightweight model are never evaluated on their own. read the letter →

arxiv 2608.01885 v1 pith:OXBX26XH submitted 2026-08-03 cs.LG

Zemin Chao , Qianhui Xu , Jianhe Cen , Guangzhi Ge , Xiao Chen , Hoangzhi Wang This is my paper
classification cs.LG
keywords timeseriesanomalydetectioncascadedinferencegatingnetworkreconstruction-basedmodel-agnosticaccelerationconfidence-guidedroutingstructureattentionnormalityconfidence
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

CARE is a cascaded inference framework for time series anomaly detection that pairs a fast, lightweight pre-filter model (LPM) with a slow, accurate complex detection model (CDM). The LPM flags high-confidence normal windows using a residual MLP autoencoder and a normality-conditioned gating network; only uncertain windows are sent to the CDM. The paper claims this routing cuts inference time by 2.7x to 4.8x relative to the most accurate existing methods, while preserving or even slightly improving detection quality on eight real-world benchmarks. The key is that most time series data is normal and predictable, so a cheap model can safely handle the bulk of traffic.

What carries the argument

The Normality-Conditioned Gating (NCG) module is the decision core: it fuses window-level, point-level, and structure-attention reconstruction features, then outputs a normality confidence score. A learnable channel-attention vector softmax-scales per-channel reconstruction deviations so that channels most indicative of anomalies dominate the gating signal. The gating network is optimized with a bipartite margin loss, a boundary hinge loss, and a confidence penalty, all anchored to a quantile of the complex model's training scores.

What would settle it

A dataset where normal patterns are highly heterogeneous and some normal windows have large reconstruction errors would break the proxy assumption. Concretely: if on such a dataset one measures the recall of the filtered-out set against the CDM's own anomaly scores, and finds that a substantial fraction of true anomalies are also filtered out (or that normal windows are routed to the CDM, negating the speedup), the core claim collapses.

Watch

Extended reading notes

Core claim

The central claim is that the uniform application of a complex anomaly detector to every time step is wasteful, because the overwhelming majority of windows are ordinary and can be judged by a simple model without loss of reliability. CARE learns a gating function that outputs a normality confidence P(x_t); samples above a quantile threshold Q_tau are scored directly by the lightweight residual MLP autoencoder, and the rest go to the complex detector. The gating network is trained with a quantile-anchored margin loss that separates low-risk from high-risk windows using the CDM's own scores as a ranking oracle, plus a structure-attention term that weights channel-wise reconstruction errors. A

Load-bearing premise

For samples the gating network filters out, the final anomaly score is just the lightweight model's reconstruction error at the current time step, so the quality guarantee rests on the assumption that this reconstruction error is a faithful proxy for how anomalous a sample is across the diverse normal patterns in the data.

Editorial extensions

If this is right

  • If CARE holds, existing high-quality but slow anomaly detectors can be deployed at a fraction of their original inference cost, making them practical for latency-sensitive or resource-constrained settings.
  • The model-agnostic design means any future high-capacity detector can be dropped into the CDM slot and immediately inherit the speedup, as demonstrated with CrossAD, MLPMixer, TranAD, and ATrans backends.
  • The pre-filtering step can reduce false positives by preventing the complex model from over-interpreting normal patterns, which the paper shows on KDD21 visualizations.
  • The two-stage training recipe (reconstruction first, then gating) is a reusable pattern for other cascaded inference problems beyond anomaly detection.
  • The MLE-based score alignment provides a principled way to mix heterogeneous anomaly scores from different models into one comparable scale.

Reading between the lines

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

  • A natural extension is to make the filtering threshold tau adaptive per stream or per time segment, since the paper itself notes the optimal tau varies with data distribution; a learned threshold could improve robustness under drift.
  • The structure-attention weights could be interpreted as a channel-importance diagnostic, offering an inexpensive byproduct for root-cause analysis when an anomaly is flagged.
  • The cascaded idea could be tested in an online or streaming setting where the gating model is updated incrementally, though the current framework assumes a fixed pre-trained CDM and static threshold.
  • One testable extension is to apply the same confidence-gated routing to other reconstruction-based detectors where the simple model is not an MLP but, say, a linear Gaussian model, to see if speedup persists without the residual MLP's inductive bias.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes CARE, a cascaded inference framework for time series anomaly detection. A lightweight Residual MLP AutoEncoder (RMA) plus a Normality-Conditioned Gating (NCG) network decides for each sliding window whether it is a high-confidence normal sample that can be scored by the cheap LPM (using the point-level reconstruction error f_pot) or whether it must be sent to a complex detector (CDM) for refined scoring. NCG is trained with the CDM's training-set anomaly scores as pseudo-labels, using a quantile-anchored margin loss with bipartite and boundary terms. Experiments on eight datasets and 14 baselines report 2.7x–4.8x speedups over the strongest baseline while maintaining or slightly improving detection quality, plus ablations, hyperparameter sensitivity, and adaptability to four different CDM backbones.

Significance. If the claims hold, CARE addresses a practically important bottleneck: deep TSAD models are expensive and anomalies are rare, so a reliable cascade is a sensible way to cut inference cost. The empirical evaluation is unusually broad (8 datasets, 14 baselines, multiple backbones, timing decompositions, stability runs), and the paper gives concrete pseudocode (Algorithm 1), full experimental tables, and a transparent description of hyperparameters. The two-stage training idea and the use of CDM score rankings to supervise the gating network are reasonable and not circular. The main risk is that the quality of the filtered-subset scores is not directly established; the paper's own GECCO RMA-routing ablation suggests that the RMA score can be misleading on some data, and this is precisely the score used for the majority of test samples.

major comments (3)
  1. [Algorithm 1 / Eq. (7) / Table 2] Algorithm 1 (line 25) sets AS_l = f_pot, and Eq. (7) uses this score for every window routed to the high-confidence normal set N. For the roughly 80% of test samples that never reach the CDM, detection quality therefore depends entirely on the RMA's point-level reconstruction error. Table 2 shows that on GECCO, replacing NCG with RMA-based routing collapses Aff-F from 0.873 (random routing) to 0.514, i.e., the RMA score is not merely noisy but actively misleading on that dataset. In full CARE the NCG mask avoids that particular failure, but the final scores on N are still f_pot, and no experiment isolates the quality of these LPM scores on the actual filtered set. The aggregate results could in principle be carried by the ~20% CDM-routed anomalies. Please report per-subset (N vs. U) Aff-F/A-P, or an ablation that scores all test points with f_pot and compares with the full CDM, and expla
  2. [Table 1 / Appendix I] The main quality table reports a single run per method with no variance or significance information. Several decisive comparisons are small (e.g., CARE 0.694 vs. DCdetector 0.690 on NYC; average Aff-F 0.781 vs. CrossAD 0.769). Appendix I reports standard deviations only for CARE on four datasets, not for the baselines. Given the central claim is 'maintaining competitive detection quality,' the authors should provide mean ± std over multiple seeds for all methods or a paired significance test (e.g., Wilcoxon signed-rank) for the quality differences.
  3. [Section 4.4 / Table 3] The paper claims model-agnostic adaptability, but it does not state whether the LPM and NCG are retrained from scratch for each CDM backbone. Since NCG training uses the CDM's training-set anomaly scores as supervision (Section 3.4), an LPM trained for one backend is not guaranteed to transfer to another. Please clarify the training protocol (retraining yes/no, threshold re-tuning per backbone) and, if no retraining, justify why the comparison is valid.
minor comments (6)
  1. [Abstract / Introduction] The claim 'the first cascaded inference framework tailored for time series anomaly detection' is a strong novelty assertion; please soften it or cite prior cascade designs for TSAD if any exist.
  2. [Figure 3] The x-axis tick labels in Figure 3 appear garbled (repeated character-code sequences); the figure needs regeneration for legibility.
  3. [Section 4 / References] The baseline list cites CATCH as [10], but reference [10] is TSINR; CATCH is reference [16]. Please correct the citation.
  4. [Appendix K] The limitation that a unified filtering threshold τ is used across all datasets is only stated in the appendix; it should be mentioned in the main text since τ is the main efficiency/quality knob.
  5. [Eq. (2) / Algorithm 1] The notation f_pot is defined through ChannelPool over the last window position, but Algorithm 1 line 25 sets AS_l = f_pot without spelling out this temporal indexing; make the definition of the point-level score earlier and more prominent.
  6. [Appendix I] The stability analysis reports only CARE; adding a comparison for the baselines under the same seeded protocol would strengthen the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: CARE's routing is trained on CDM training scores as distillation, and test quality is evaluated on external benchmarks.

full rationale

The paper's derivation chain is self-contained and does not reduce to its own inputs. The LPM (RMA + NCG) is trained in two stages: RMA is trained solely by reconstruction loss (Eq. 3), and NCG is trained by ranking losses (Eqs. 4-6) using the CDM's training-set anomaly scores as pseudo-labels. This is a distillation-like routing objective, not a claim that CARE's output is derived from its own final predictions. At inference, high-confidence samples receive the LPM's point-level reconstruction error f_pot (Algorithm 1, line 25) and uncertain samples receive the CDM score, with the two score spaces aligned via log-normal MLE CDFs (Eq. 7). All fitted parameters — RMA weights, NCG weights, and alignment parameters — are estimated from training data only. The reported detection-quality metrics (Aff-F, A-P, etc.) are computed on held-out test labels against 14 external baselines; no test labels or metrics are used as training inputs. The GECCO RMA-routing ablation (Table 2) shows that f_pot can be an unreliable score for some datasets, but this is a correctness/robustness concern about the filtered-subset score, not a circularity: the final score for filtered samples is f_pot by architectural design, yet the claim that CARE 'maintains competitive detection quality' is an empirical claim tested on benchmarks, not an entailment of the training objective. The paper invokes no uniqueness theorems, no load-bearing self-citations, and no ansatz smuggled in via prior work by the same authors. The only self-referential element is that the CDM used for supervision is also used for uncertain test samples, which is a standard teacher-student / cascade design and does not by construction force the reported accuracy or speedup. Therefore the paper receives a circularity score of 0.

Assumptions & free parameters 8 free parameters · 4 assumptions · 0 invented entities

The framework's central empirical claim depends on a set of hand-chosen hyperparameters (thresholds, margins, weights), a domain assumption that reconstruction error is a valid anomaly score, and the availability of a teacher CDM. None of these are derived from first principles; they are standard choices for a learned router. The log-normal alignment assumption is shown empirically to be non-critical (Gaussian and empirical CDF variants are competitive).

free parameters (8)
  • Filtering threshold tau = 0.2
    Controls the fraction of samples routed to CDM; chosen by hand with sensitivity analysis, not dataset-specific.
  • Risk quantile Q_risk = 0.7
    Quantile of CDM training scores partitioning safe vs risky in NCG training.
  • Boundary width w = 0.1
    Width of the hinge band in the boundary loss.
  • Margin hyperparameters xi1, xi2 = 0.01
    Margins in the bipartite and boundary losses.
  • Regularization weights alpha, lambda = alpha=0.05, lambda=1.0
    Weights for the boundary loss and confidence penalty in Eq. (6).
  • Window size L = per-dataset, matched to CDM
    Sliding window length; set to the CDM's window size for semantic consistency.
  • Hidden dimensions Dh, Dz, Dg = 16, 5, 5
    Architecture sizes for RMA and NCG.
  • Channel attention vector v = learned, no fixed value
    Learnable softmax weights in structure attention, fitted during training.
assumptions (4)
  • domain assumption The CDM provides deterministic continuous anomaly scores
    Stated in Section 3.1: CARE assumes the backend CDM outputs a continuous anomaly score, which is standard for TSAD models.
  • domain assumption Anomaly scores are log-normally distributed
    Section 3.4 and Appendix F fit log-normal distributions via MLE for score alignment; Table 5 shows the method is robust to this choice, so it is not critical.
  • domain assumption Reconstruction error of a lightweight autoencoder trained on normal data is a valid anomaly severity signal for filtered samples
    The final score for filtered samples is AS_l = f_pot, the per-channel reconstruction error at the current time step (Algorithm 1, NCG procedure). The quality guarantee depends on this error being low for true normals and high for anomalies.
  • domain assumption The training-set CDM scores are a reliable teacher for routing decisions
    NCG is supervised with ASc on the training set using a quantile threshold (Section 3.4, Eqs. 4-5); if these scores are miscalibrated, the router inherits the error.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CARE: A Cascaded Framework for Efficient and Reliable Time Series Anomaly Detection." pith.science (2026). https://pith.science/paper/OXBX26XH

@misc{pith2026260801885,
  author       = {Pith},
  title        = {Pith review of: CARE: A Cascaded Framework for Efficient and Reliable Time Series Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OXBX26XH}},
  note         = {Machine review of arXiv:2608.01885}
}
abstract

While deep learning models have achieved state-of-the-art performance in time series anomaly detection, their complex architectures incur substantial inference overhead. Existing methods typically apply a uniform inference strategy across all data points, which is inefficient given that anomalies are inherently scarce and the vast majority of temporal data consists of predictable normal patterns. To mitigate this bottleneck, we propose CARE, a model-agnostic cascaded inference framework that integrates a Lightweight Pre-filter Model (LPM) with an existing high-capacity Complex Detection Model (CDM). The LPM rapidly filters high-confidence normal samples using a Residual MLP AutoEncoder and a Normality-Conditioned Gating mechanism. Crucially, we introduce a Structure Attention module to explicitly capture channel-wise anomaly contributions, and optimize the gating network via a confidence-guided selective routing objective that learns reliable routing decisions to reduce unnecessary CDM invocations. Extensive experiments across eight real-world benchmarks demonstrate that CARE effectively isolates high-confidence normal samples. By routing only uncertain samples to the CDM, our framework achieves $2.7\times$ to $4.8\times$ inference speedup compared to the most accurate SOTA approaches, while still maintaining competitive detection quality.

Figures

Figures reproduced from arXiv: 2608.01885 by the authors.

Figure 1
Figure 1. Motivation of the proposed CARE framework. (a) The trade-off between detection [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall architecture of the proposed CARE framework. (a) The Residual AutoEncoder [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Parameter sensitivity studies of main hyper-parameters in CARE. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Inference time comparison of CARE with various CDM backbones. [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Visualization of anomalies detected by CARE and CrossAD on the KDD21 dataset. [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: Empirical distributions of training anomaly scores and their fitted log-normal PDFs on [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]
Figure 7
Figure 7. Figure 7: Additional parameter sensitivity studies of main hyperparameters in CARE. [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: Additional inference time comparison of CARE with various CDM backbones. [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

51 extracted references · 3 linked inside Pith

  1. [1]

    Deep learning for time series anomaly detection: A survey.ACM Computing Surveys, 57(1):1– 42, 2024

    Zahra Zamanzadeh Darban, Geoffrey I Webb, Shirui Pan, Charu Aggarwal, and Mahsa Salehi. Deep learning for time series anomaly detection: A survey.ACM Computing Surveys, 57(1):1– 42, 2024

    2024

  2. [2]

    D. M. Hawkins.Identification of Outliers. Monographs on Applied Probability and Statistics. Springer, 1980

    1980

  3. [3]

    A review on out- lier/anomaly detection in time series data.ACM computing surveys (CSUR), 54(3):1–33, 2021

    Ane Blázquez-García, Angel Conde, Usue Mori, and Jose A Lozano. A review on out- lier/anomaly detection in time series data.ACM computing surveys (CSUR), 54(3):1–33, 2021

    2021

  4. [4]

    Credit card fraud detection using advanced transformer model

    Chang Yu, Yongshun Xu, Jin Cao, Ye Zhang, Yixin Jin, and Mengran Zhu. Credit card fraud detection using advanced transformer model. In2024 IEEE international conference on metaverse computing, networking, and applications (MetaCom), pages 343–350. IEEE, 2024

    2024

  5. [5]

    Ecg-net: A deep lstm autoencoder for detecting anomalous ecg.Engineering Applications of Artificial Intelligence, 124:106484, 2023

    Moumita Roy, Sukanta Majumder, Anindya Halder, and Utpal Biswas. Ecg-net: A deep lstm autoencoder for detecting anomalous ecg.Engineering Applications of Artificial Intelligence, 124:106484, 2023

    2023

  6. [6]

    Anomaly detection in event- triggered traffic time series via similarity learning.IEEE Transactions on Dependable and Secure Computing, 22(2):888–902, 2024

    Shaoyu Dou, Kai Yang, Yang Jiao, Chengbo Qiu, and Kui Ren. Anomaly detection in event- triggered traffic time series via similarity learning.IEEE Transactions on Dependable and Secure Computing, 22(2):888–902, 2024

    2024

  7. [7]

    Lof: identifying density-based local outliers

    Markus M Breunig, Hans-Peter Kriegel, Raymond T Ng, and Jörg Sander. Lof: identifying density-based local outliers. InProceedings of the 2000 ACM SIGMOD international conference on Management of data, pages 93–104, 2000

    2000

  8. [8]

    Isolation forest

    Fei Tony Liu, Kai Ming Ting, and Zhi-Hua Zhou. Isolation forest. In2008 eighth ieee international conference on data mining, pages 413–422. IEEE, 2008

    2008

Show all 51 references
  1. [9]

    Support vector method for novelty detection.Advances in neural information processing systems, 12, 1999

    Bernhard Schölkopf, Robert C Williamson, Alex Smola, John Shawe-Taylor, and John Platt. Support vector method for novelty detection.Advances in neural information processing systems, 12, 1999

    1999

  2. [10]

    Tsinr: capturing temporal continuity via implicit neural representations for time series anomaly detection

    Mengxuan Li, Ke Liu, Hongyang Chen, Jiajun Bu, Hongwei Wang, and Haishuai Wang. Tsinr: capturing temporal continuity via implicit neural representations for time series anomaly detection. InProceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V ...

    2025

  3. [11]

    Sarad: Spatial association- aware anomaly detection and diagnosis for multivariate time series.Advances in Neural Information Processing Systems, 37:48371–48410, 2024

    Zhihao Dai, Ligang He, Shuang-Hua Yang, and Matthew Leeke. Sarad: Spatial association- aware anomaly detection and diagnosis for multivariate time series.Advances in Neural Information Processing Systems, 37:48371–48410, 2024

    2024

  4. [12]

    Crossad: Time series anomaly detection with cross-scale associations and cross-window modeling.arXiv preprint arXiv:2510.12489, 2025

    Beibu Li, Qichao Shentu, Yang Shu, Hui Zhang, Ming Li, Ning Jin, Bin Yang, and Chenjuan Guo. Crossad: Time series anomaly detection with cross-scale associations and cross-window modeling.arXiv preprint arXiv:2510.12489, 2025

    2025

  5. [13]

    Mlp-mixer based masked autoen- coders are effective, explainable and robust for time series anomaly detection.Proceedings of the VLDB Endowment, 18(3):798–811, 2024

    Qideng Tang, Chaofan Dai, Yahui Wu, and Haohao Zhou. Mlp-mixer based masked autoen- coders are effective, explainable and robust for time series anomaly detection.Proceedings of the VLDB Endowment, 18(3):798–811, 2024

    2024

  6. [14]

    Noise matters: Cross contrastive learning for flink anomaly detection.Proceedings of the VLDB Endowment, 18(4):1159–1168, 2024

    Zhihao Zhuang, Yingying Zhang, Kai Zhao, Chenjuan Guo, Bin Yang, Qingsong Wen, and Lunting Fan. Noise matters: Cross contrastive learning for flink anomaly detection.Proceedings of the VLDB Endowment, 18(4):1159–1168, 2024

    2024

  7. [15]

    Temporal- frequency masked autoencoders for time series anomaly detection

    Yuchen Fang, Jiandong Xie, Yan Zhao, Lu Chen, Yunjun Gao, and Kai Zheng. Temporal- frequency masked autoencoders for time series anomaly detection. In2024 IEEE 40th interna- tional conference on data engineering (ICDE), pages 1228–1241. IEEE, 2024

    2024

  8. [16]

    Catch: Channel-aware multivariate time series anomaly detection via frequency patching.arXiv preprint arXiv:2410.12261, 2024

    Xingjian Wu, Xiangfei Qiu, Zhengyu Li, Yihang Wang, Jilin Hu, Chenjuan Guo, Hui Xiong, and Bin Yang. Catch: Channel-aware multivariate time series anomaly detection via frequency patching.arXiv preprint arXiv:2410.12261, 2024. 10

    2024 arXiv

  9. [17]

    Tab: Unified benchmarking of time series anomaly detection methods.Proceedings of the VLDB Endowment, 18(9):2775–2789, 2025

    Xiangfei Qiu, Zhe Li, Wanghui Qiu, Shiyan Hu, Lekui Zhou, Xingjian Wu, Zhengyu Li, Chenjuan Guo, Aoying Zhou, Zhenli Sheng, et al. Tab: Unified benchmarking of time series anomaly detection methods.Proceedings of the VLDB Endowment, 18(9):2775–2789, 2025

    2025

  10. [18]

    An introduction to outlier analysis

    Charu C Aggarwal. An introduction to outlier analysis. InOutlier analysis, pages 1–34. Springer, 2016

    2016

  11. [19]

    Anomaly detection in time series: a comprehensive evaluation.Proceedings of the VLDB Endowment, 15(9):1779–1797, 2022

    Sebastian Schmidl, Phillip Wenig, and Thorsten Papenbrock. Anomaly detection in time series: a comprehensive evaluation.Proceedings of the VLDB Endowment, 15(9):1779–1797, 2022

    2022

  12. [20]

    itransformer: Inverted transformers are effective for time series forecasting.The Twelfth International Conference on Learning Representations (ICLR), 2024

    Yong Liu, Tengge Hu, Haoran Zhang, Haixu Wu, Shiyu Wang, Lintao Ma, and Mingsheng Long. itransformer: Inverted transformers are effective for time series forecasting.The Twelfth International Conference on Learning Representations (ICLR), 2024

    2024

  13. [21]

    One fits all: Power general time series analysis by pretrained lm.Advances in neural information processing systems, 36:43322–43355, 2023

    Tian Zhou, Peisong Niu, Liang Sun, Rong Jin, et al. One fits all: Power general time series analysis by pretrained lm.Advances in neural information processing systems, 36:43322–43355, 2023

    2023

  14. [22]

    Dynamic neural networks: A survey.IEEE transactions on pattern analysis and machine intelligence, 44(11):7436–7456, 2021

    Yizeng Han, Gao Huang, Shiji Song, Le Yang, Honghui Wang, and Yulin Wang. Dynamic neural networks: A survey.IEEE transactions on pattern analysis and machine intelligence, 44(11):7436–7456, 2021

    2021

  15. [23]

    Frugalgpt: How to use large language models while reducing cost and improving performance.Trans

    Lingjiao Chen, Matei Zaharia, and James Zou. Frugalgpt: How to use large language models while reducing cost and improving performance.Trans. Mach. Learn. Res., 2024

    2024

  16. [24]

    Tranad: Deep transformer networks for anomaly detection in multivariate time series data.Proc

    Shreshth Tuli, Giuliano Casale, and Nicholas R Jennings. Tranad: Deep transformer networks for anomaly detection in multivariate time series data.Proc. VLDB Endow., 15(6):1201–1214, 2022

    2022

  17. [25]

    Uci machine learning repository, 2007

    Arthur Asuncion and David Newman. Uci machine learning repository, 2007

    2007

  18. [26]

    A comparative study of htm and other neural network models for online sequence learning with streaming data

    Yuwei Cui, Chetan Surpur, Subutai Ahmad, and Jeff Hawkins. A comparative study of htm and other neural network models for online sequence learning with streaming data. In2016 International joint conference on neural networks (IJCNN), pages 1530–1538. IEEE, 2016

    2016

  19. [27]

    Gecco industrial challenge 2018 dataset.Tech

    S Moritz, F Rehbach, S Chandrasekaran, M Rebolledo, and T Bartz-Beielstein. Gecco industrial challenge 2018 dataset.Tech. Rep., 2018

    2018

  20. [28]

    Detecting spacecraft anomalies using lstms and nonparametric dynamic thresholding

    Kyle Hundman, Valentino Constantinou, Christopher Laporte, Ian Colwell, and Tom Soderstrom. Detecting spacecraft anomalies using lstms and nonparametric dynamic thresholding. In Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining, p...

    2018

  21. [29]

    Practical approach to asynchronous multivariate time series anomaly detection and localization

    Ahmed Abdulaal, Zhuanghua Liu, and Tomer Lancewicki. Practical approach to asynchronous multivariate time series anomaly detection and localization. InProceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining, pages 2485–2494, 2021

    2021

  22. [30]

    Ultrafast local outlier detection from a data stream with stationary region skipping

    Susik Yoon, Jae-Gil Lee, and Byung Suk Lee. Ultrafast local outlier detection from a data stream with stationary region skipping. InProceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 1181–1191, 2020

    2020

  23. [31]

    Merlion: A machine learning library for time series.arXiv preprint arXiv:2109.09265, 2021

    Aadyot Bhatnagar, Paul Kassianik, Chenghao Liu, Tian Lan, Wenzhuo Yang, Rowan Cassius, Doyen Sahoo, Devansh Arpit, Sri Subramanian, Gerald Woo, et al. Merlion: A machine learning library for time series.arXiv preprint arXiv:2109.09265, 2021

    2021 arXiv

  24. [32]

    Anomaly detection using autoencoders with nonlinear dimensionality reduction

    Mayu Sakurada and Takehisa Yairi. Anomaly detection using autoencoders with nonlinear dimensionality reduction. InProceedings of the MLSDA 2014 2nd workshop on machine learning for sensory data analysis, pages 4–11, 2014

    2014

  25. [33]

    Anomaly transformer: Time series anomaly detection with association discrepancy

    Jiehui Xu, Haixu Wu, Jianmin Wang, and Mingsheng Long. Anomaly transformer: Time series anomaly detection with association discrepancy. InThe Tenth International Conference on Learning Representations (ICLR). OpenReview.net, 2022. 11

    2022

  26. [34]

    Dcdetector: Dual attention contrastive representation learning for time series anomaly detection

    Yiyuan Yang, Chaoli Zhang, Tian Zhou, Qingsong Wen, and Liang Sun. Dcdetector: Dual attention contrastive representation learning for time series anomaly detection. InProceedings of the 29th ACM SIGKDD conference on knowledge discovery and data mining, pages 3033–3045, 2023

    2023

  27. [35]

    Timesnet: Temporal 2d-variation modeling for general time series analysis

    Haixu Wu, Tengge Hu, Yong Liu, Hang Zhou, Jianmin Wang, and Mingsheng Long. Timesnet: Temporal 2d-variation modeling for general time series analysis. InThe Eleventh International Conference on Learning Representations (ICLR). OpenReview.net, 2023

    2023

  28. [36]

    A time series is worth 64 words: Long-term forecasting with transformers

    Yuqi Nie, Nam H Nguyen, Phanwadee Sinthong, and Jayant Kalagnanam. A time series is worth 64 words: Long-term forecasting with transformers. InThe Eleventh International Conference on Learning Representations (ICLR). OpenReview.net, 2023

    2023

  29. [37]

    Calf: Aligning llms for time series forecasting via cross-modal fine-tuning

    Peiyuan Liu, Hang Guo, Tao Dai, Naiqi Li, Jigang Bao, Xudong Ren, Yong Jiang, and Shu-Tao Xia. Calf: Aligning llms for time series forecasting via cross-modal fine-tuning. InProceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 18915–18923, 2025

    2025

  30. [38]

    Graph neural network-based anomaly detection in multivariate time series

    Ailin Deng and Bryan Hooi. Graph neural network-based anomaly detection in multivariate time series. InProceedings of the AAAI conference on artificial intelligence, volume 35, pages 4027–4035, 2021

    2021

  31. [39]

    Beyond sharing: Conflict-aware multivariate time series anomaly detection

    Haotian Si, Changhua Pei, Zhihan Li, Yadong Zhao, Jingjing Li, Haiming Zhang, Zulong Diao, Jianhui Li, Gaogang Xie, and Dan Pei. Beyond sharing: Conflict-aware multivariate time series anomaly detection. InProceedings of the 31st ACM Joint European Software Engineering Confere...

    2023

  32. [40]

    Robust anomaly detection for multivariate time series through stochastic recurrent neural network

    Ya Su, Youjian Zhao, Chenhao Niu, Rong Liu, Wei Sun, and Dan Pei. Robust anomaly detection for multivariate time series through stochastic recurrent neural network. InProceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, pages 2828–2...

    2019

  33. [41]

    Usad: Unsupervised anomaly detection on multivariate time series

    Julien Audibert, Pietro Michiardi, Frédéric Guyard, Sébastien Marti, and Maria A Zuluaga. Usad: Unsupervised anomaly detection on multivariate time series. InProceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 3395–3404, 2020

    2020

  34. [42]

    Are transformers effective for time series forecasting? InProceedings of the AAAI conference on artificial intelligence, volume 37, pages 11121–11128, 2023

    Ailing Zeng, Muxi Chen, Lei Zhang, and Qiang Xu. Are transformers effective for time series forecasting? InProceedings of the AAAI conference on artificial intelligence, volume 37, pages 11121–11128, 2023

    2023

  35. [43]

    Timexer: Empowering transformers for time series forecasting with exogenous variables.Advances in Neural Information Processing Systems, 37:469–498, 2024

    Yuxuan Wang, Haixu Wu, Jiaxiang Dong, Guo Qin, Haoran Zhang, Yong Liu, Yunzhong Qiu, Jianmin Wang, and Mingsheng Long. Timexer: Empowering transformers for time series forecasting with exogenous variables.Advances in Neural Information Processing Systems, 37:469–498, 2024

    2024

  36. [44]

    Robust real-time face detection.International journal of computer vision, 57(2):137–154, 2004

    Paul Viola and Michael J Jones. Robust real-time face detection.International journal of computer vision, 57(2):137–154, 2004

    2004

  37. [45]

    Cascade r-cnn: Delving into high quality object detection

    Zhaowei Cai and Nuno Vasconcelos. Cascade r-cnn: Delving into high quality object detection. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 6154–6162, 2018

    2018

  38. [46]

    Make your vit-based multi-view 3d detectors faster via token compression

    Dingyuan Zhang, Dingkang Liang, Zichang Tan, Xiaoqing Ye, Cheng Zhang, Jingdong Wang, and Xiang Bai. Make your vit-based multi-view 3d detectors faster via token compression. In European Conference on Computer Vision, pages 56–72. Springer, 2024

    2024

  39. [47]

    Routerdc: Query-based router by dual contrastive learning for assembling large language models.Advances in Neural Information Processing Systems, 37:66305–66328, 2024

    Shuhao Chen, Weisen Jiang, Baijiong Lin, James Kwok, and Yu Zhang. Routerdc: Query-based router by dual contrastive learning for assembling large language models.Advances in Neural Information Processing Systems, 37:66305–66328, 2024

    2024

  40. [48]

    Cascadia: A cascade serving system for large language models.arXiv e-prints, pages arXiv–2506, 2025

    Youhe Jiang, Fangcheng Fu, Wanru Zhao, Stephan Rabanser, Nicholas D Lane, and Binhang Yuan. Cascadia: A cascade serving system for large language models.arXiv e-prints, pages arXiv–2506, 2025. 12

    2025

  41. [49]

    A survey on efficient inference for large language models.arXiv preprint arXiv:2404.14294, 2024

    Zixuan Zhou, Xuefei Ning, Ke Hong, Tianyu Fu, Jiaming Xu, Shiyao Li, Yuming Lou, Luning Wang, Zhihang Yuan, Xiuhong Li, et al. A survey on efficient inference for large language models.arXiv preprint arXiv:2404.14294, 2024

    2024 arXiv

  42. [50]

    Data shunt: Collaboration of small and large models for lower costs and better performance

    Dong Chen, Yueting Zhuang, Shuo Zhang, Jinfeng Liu, Su Dong, and Siliang Tang. Data shunt: Collaboration of small and large models for lower costs and better performance. InProceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 11249–11257, 2024

    2024

  43. [51]

    A novel anomaly detection scheme based on principal component classifier

    Mei-Ling Shyu, Shu-Ching Chen, Kanoksri Sarinnapakorn, and LiWu Chang. A novel anomaly detection scheme based on principal component classifier. 2003. 13 A Extended Related Work A.1 Deep Time Series Anomaly Detection Deep learning-based TSAD utilizes neural networks to learn f...

    2003

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.