Pith. sign in

REVIEW 4 major objections 5 minor 52 references

SenDaL: An Effective and Efficient Calibration Framework of Low-Cost Sensors for Daily Life

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

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

Pith's one-line read SenDaL is a neural-network calibration framework that routes each low-cost sensor reading either to a fast linear layer or to a deep model, and the paper reports it keeps deep-model accuracy while running at near-linear speed.

desk verdict Practical adaptive-inference framework with a thorough hardware evaluation, but the real-time accuracy claim rests on a possibly non-causal HP filter that the paper never discusses. read the letter →

arxiv 2502.10460 v1 pith:5IJAPCIM submitted 2025-02-12 cs.LG

Seokho Ahn , Hyungjin Kim , Euijong Lee , Young-Duk Seo This is my paper · ORCID
classification cs.LG
keywords low-costsensorcalibrationIoTmodelroutingbottom-uptrainingtop-downinferencetime-seriesLSTMTransformer
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

SenDaL is a neural-network calibration framework for low-cost sensors embedded in IoT devices. It is built on the observation that sensor readings are stable most of the time, with sharp changes only occasionally. A small trained classifier decides, window by window, whether to correct the reading with a cheap linear layer or with a deep model such as an LSTM or Transformer; because the cheap path handles most windows, the framework reports accuracy close to deep models while keeping latency and energy close to a linear model. The paper verifies this on fine-dust sensors in three real indoor environments and on embedded hardware including Raspberry Pi and Arduino. If the claims hold, everyday home appliances could field deep-learning-grade sensor calibration on their existing CPUs.

What carries the argument

The load-bearing object is the embedding layer $E = \sigma \circ L$: a linear layer followed by a sigmoid that classifies each time-series window as stable or unstable. Training proceeds bottom-up: first the linear layer $f_c$ and the component layer $C$ are trained independently; then soft class labels $y^*_i$ are built from their relative errors, a smoothed-hard label, an instability factor, and a magnifier, and $E$ is trained by binary cross-entropy; finally all layers are fine-tuned together with $E$ frozen, using threshold $\theta = 0.5$ to route each window. At inference the framework applies $F(S) = C(S)$ if $E(S) > \theta$ and $F(S) = f_c(S)$ otherwise, and it may skip evaluating $E$ on every window when nearby windows are stable. This routing mechanism is what converts deep-model accuracy into linear-model cost.

What would settle it

Re-run the Env1-Env3 experiments with a strictly causal reference: replace the two-sided Hodrick-Prescott filter with a one-sided or online filter, and compare SenDaL's RMSE and miss ratio against the linear baseline. If the accuracy gap disappears or shrinks to the level of the baselines, the router's reported gain came from future information baked into the labels rather than from the routing mechanism itself.

Watch

Extended reading notes

Core claim

The paper's central claim is that you do not need to run an expensive deep model on every sensor reading to get deep-model accuracy. SenDaL trains a deep model and a linear model side by side, then trains a third component — an embedding layer with a sigmoid output — to predict which of the two will do better on the current window of readings. During inference the embedding layer alone chooses the path: stable windows go through the linear layer, unstable windows go through the deep component. Because stable windows dominate in daily-life data, average inference cost lands near the linear model while calibration error stays at the deep-model level; the miss-ratio results indicate the framework also reduces the large errors to which plain linear calibration is prone.

Load-bearing premise

The evaluation treats the smoothed average of eight laser fine-dust sensors as the correct value the framework must reproduce; if that reference signal is not actually available in real time — for example, because the smoothing filter uses future readings — the reported accuracy improvement would not carry over to live deployment.

Editorial extensions

If this is right

  • On CPU-only embedded boards, SenDaL runs an order of magnitude faster than the corresponding deep model: roughly 7x faster with GRU and 20x faster with Phased LSTM on the reported Raspberry Pi measurements.
  • Because the embedding layer can be paired with any component model, the same framework can trade accuracy for speed by swapping LSTM, GRU, Transformer, Reformer, or Performer as hardware resources change.
  • The miss-ratio results say the framework does not merely match average error; it cuts the rate of large calibration errors relative to linear regression, e.g. a 43.5% improvement at threshold 30 in the reported experiments.
  • The trained embedding layer gives a running stability signal for the environment, so the framework can flag unstable periods that deserve the more expensive correction path.
  • ONNX and FlatBuffers conversions keep the speed advantage on microcontrollers such as Arduino Nano, widening the range of devices that can host it.

Reading between the lines

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

  • Beyond the paper, the same train-a-router-on-two-teachers recipe should transfer to other cheap-sensor pairs, such as temperature, humidity, gas, or air-pressure sensors, where a stable baseline is common and a high-cost reference is available only during training; SenDaL itself is demonstrated only on fine dust.
  • One could add an explicit budget to the classifier so the fraction of deep-path calls is capped, turning the accuracy-cost trade into a user-set knob; the paper fixes the threshold at 0.5 and does not explore this.
  • If the locality assumption justifies skipping the embedding layer between stable windows, an adaptive scheduler could skip more aggressively as the environment becomes calm, further reducing cost without retraining.
  • The framework's behavior on unusual or adversarial inputs is untested: the router will classify a genuinely novel pattern as unstable and pay the deep cost, but there is no guarantee the deep component is accurate on distributions absent from training.
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

4 major / 5 minor

Summary. The paper proposes SenDaL, a gating framework for calibrating low-cost IoT sensors. It trains a linear calibration layer fc and a deeper 'component' layer C (LSTM, GRU, Transformer, etc.) separately, then trains an embedding classifier E on soft labels derived from which layer achieved lower error on each training window. At inference, E selects one of the two layers, so the deep model is only invoked on windows classified as unstable. The authors evaluate on fine-dust data collected with PPD42NS infrared sensors and PMS7003 laser sensors in three environments, reporting RMSE, miss ratio, CPU latency, and energy consumption on desktop CPUs, Raspberry Pi 3/4, Jetson Nano, and Arduino Nano. They report that SenDaL approaches or improves on the accuracy of the deep component while reducing latency and energy by large factors.

Significance. If the reported accuracy numbers hold, SenDaL is a practically useful routing scheme: it gives a concrete way to keep most inferences on a cheap linear model while reserving a deep model for unstable segments, and the paper's hardware evaluation is unusually thorough (multiple boards, PyTorch/ONNX/FlatBuffers formats, miss-ratio analysis). The central accuracy evidence, however, depends on preprocessing choices that are not fully specified (notably the direction of the HP filter), and several router hyperparameters are not reported. The contribution is therefore significant but conditional.

major comments (4)
  1. [Section IV-A, Eq. (2), Table II] The manuscript does not state whether the Hodrick-Prescott filter used to construct the ground truth y is causal. The standard HP filter solves a global smoothing problem over the whole series, so the smoothed value at time i depends on future observations; the same is true of the linear interpolation step if it uses neighboring future samples. If the standard two-sided filter was used, then the targets y_i in Eqs. (2)-(4) and in the RMSE of Table II contain information unavailable to F(S_i) in Eq. (9), and the reported absolute real-time accuracy is not attainable. Please specify the exact filter implementation, state whether it is one-sided or two-sided, and re-run the accuracy evaluation with a strictly causal preprocessing pipeline, or provide a direct comparison showing the effect of using causal targets.
  2. [Section V-B, Section VI] The ground truth is the average of eight PMS7003 sensors, which are themselves low-cost consumer-grade laser sensors, not a reference instrument. Section VI acknowledges this, but the abstract and introduction claim that SenDaL produces data 'accurate' and 'close to high-cost sensors'; the experiments only support accuracy relative to this proxy. The authors should either add a validation against a reference-grade instrument (e.g., a beta-ray monitor) for at least one environment, or soften the absolute-accuracy claims to claims about agreement with the multi-sensor laser average.
  3. [Section IV-D, Section V-E] Several free parameters control the router and therefore the accuracy/latency tradeoff: the margin xi in Eq. (4), the smoothing factor lambda in Eqs. (5)-(6), the weight w and magnifier n in Eq. (3), and the threshold theta used in Eqs. (8)-(9). Section V-E reports only N and hidden sizes. Without the chosen values and a sensitivity analysis, the claim that SenDaL 'approaches or surpasses' the component accuracy across environments is not yet robustly supported, since xi and theta directly determine how often the deep model is used.
  4. [Table III] The improvement percentages in Table III do not match the listed miss ratios. For theta_M=30, the linear miss ratio is 0.033 and SenDaL's is 0.023, which is a 30.3% decrease, not the reported 43.5%; similar discrepancies appear at theta_M=3, 5, and 10. Please correct the computations or clarify the definition of 'Impv.'.
minor comments (5)
  1. [Table IV] The average latency for Performer Vanilla on PyTorch is listed as 4.118 microseconds, but the three environment values are 7.287, 6.633, and 7.434 microseconds, whose average is approximately 7.118 microseconds. Please correct the typo.
  2. [Section IV-E] The sentence 'If we assume locality... By adjusting the period through E, the inference time of E is also reduced' is vague; please specify how the evaluation period is chosen and how it affects the reported latency.
  3. [Section V-B] There is a typo: 'find-dust' should be 'fine-dust'.
  4. [Section V-D] The paper reports averages of five latency measurements without variance or confidence intervals; given the large improvement factors claimed, reporting standard deviations would strengthen the hardware results.
  5. [References] Several references (e.g., [14], [21], [22], [24]) lack complete publication venues; please complete them.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SenDaL's gating labels are ordinary supervised targets and its accuracy, latency, and energy claims are benchmarked against external baselines on held-out folds.

full rationale

SenDaL's derivation chain is self-contained and does not reduce to its own inputs. The classifier labels in Eq. (3) through Eq. (6) are generated from the empirical calibration errors of the linear and component layers on training data, which is standard supervised gating rather than a definitional identity; Eq. (8) fine-tunes the two calibration layers under the classifier's decisions, and Eq. (9) is a conditional inference rule. Accuracy, latency, and energy are then measured on held-out walk-forward folds (Section V-D) against external baselines (LSTM, GRU, Phased LSTM, Transformer, Reformer, Performer, and linear regression) on three datasets and multiple embedded hardware platforms, so the reported gains are not forced by construction. The only self-citation is [2], a general IoT survey used for context in Section I, and it is not load-bearing. The potential HP-filter causality issue in Sections IV-A and V-B is a correctness or leakage risk, not a circularity, because the ground truth is not defined in terms of SenDaL's own output. Therefore, no circular step is present.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The paper is empirical, so the ledger contains hand-set hyperparameters and domain assumptions about sensors and preprocessing rather than mathematical axioms. These are the quantities the central accuracy and efficiency claims depend on that are neither derived nor externally verified in the text.

free parameters (7)
  • Window size N = 20
    Fixed to 20 samples (a 5-minute input window); no sensitivity analysis is reported.
  • Decision margin xi
    Defined in Eq. (4) to bias routing toward the linear model; its numerical value is not reported.
  • Smoothing factor lambda
    Window length used for smoothed class labels and instability factor in Eqs. (5) and (6); value not reported.
  • Weight w in soft label
    Weighting parameter in Eq. (3) balancing the smoothed hard label and instability factor; value not reported.
  • Magnifier parameter n
    Shape parameter of the magnifier f_dot in Eq. (3); only constrained to n > 0.
  • Routing threshold theta = 0.5
    Decision threshold used in Eqs. (8) and (9) to select the linear or component path.
  • Hidden sizes for LSTM and Transformer = 16 / 8
    Model capacities chosen in Section V-E; no tuning or sensitivity analysis is provided.
assumptions (4)
  • domain assumption Averaged readings from eight PMS7003 laser sensors are a valid ground truth y.
    Section V-B sets PMS7003 as the high-accuracy reference and PPD42NS as the low-cost sensor; Section VI concedes this reference may not be accurate enough.
  • domain assumption SMA and HP filtering plus linear interpolation preserve the calibration relationship while removing sensor artifacts.
    Section IV-A applies these filters to all data before training; no analysis shows that filtering does not remove the signal being modeled or leak future information.
  • ad hoc to paper The classifier E trained on soft labels derived from fc and C errors generalizes to unseen windows.
    Eqs. (3)-(7) define the soft labels using the two calibration models on training data; top-down inference relies entirely on this generalization.
  • domain assumption Anchored walk-forward folds are independent despite temporal autocorrelation in the sensor series.
    Section V-D describes 10-fold anchored walk-forward optimization; no correction for overlapping windows or autocorrelation is described.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SenDaL: An Effective and Efficient Calibration Framework of Low-Cost Sensors for Daily Life." pith.science (2026). https://pith.science/paper/5IJAPCIM

@misc{pith2026250210460,
  author       = {Pith},
  title        = {Pith review of: SenDaL: An Effective and Efficient Calibration Framework of Low-Cost Sensors for Daily Life},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5IJAPCIM}},
  note         = {Machine review of arXiv:2502.10460}
}
read the original abstract

The collection of accurate and noise-free data is a crucial part of Internet of Things (IoT)-controlled environments. However, the data collected from various sensors in daily life often suffer from inaccuracies. Additionally, IoT-controlled devices with low-cost sensors lack sufficient hardware resources to employ conventional deep-learning models. To overcome this limitation, we propose sensors for daily life (SenDaL), the first framework that utilizes neural networks for calibrating low cost sensors. SenDaL introduces novel training and inference processes that enable it to achieve accuracy comparable to deep learning models while simultaneously preserving latency and energy consumption similar to linear models. SenDaL is first trained in a bottom-up manner, making decisions based on calibration results from both linear and deep learning models. Once both models are trained, SenDaL makes independent decisions through a top-down inference process, ensuring accuracy and inference speed. Furthermore, SenDaL can select the optimal deep learning model according to the resources of the IoT devices because it is compatible with various deep learning models, such as long short-term memory-based and Transformer-based models. We have verified that SenDaL outperforms existing deep learning models in terms of accuracy, latency, and energy efficiency through experiments conducted in different IoT environments and real-life scenarios.

Figures

Figures reproduced from arXiv: 2502.10460 by the authors.

Figure 1
Figure 1. Overview of SenDaL. Both high-cost and low-cost sensors are used to train SenDaL on a machine through a bottom-up training process. Then, a top-down inference process enables SenDaL calibrate low-cost sensors in IoT-controlled devices with high accuracy and fast inference speed. [7]–[11] have calibrated the sensor data using deep learning techniques and demonstrated remarkable performance. Despite these advancements… view at source ↗
Figure 2
Figure 2. Overall data refinement process. models are mainly used in their domains [7], [8]. Moreover, research has focused on monitoring the sensor reliability or calibrating data based on multiple low-cost sensors [19], [36], [37]. Nevertheless, these methods are not sufficiently simple for increasing the accuracy of a single low-cost sensor and do not improve the latency because they do not consider real￾time inference. In… view at source ↗
Figure 4
Figure 4. Two types of decision-making concept. (a) Embedding layer (Classifier) learns a model determination method using the results of several calibration models, in a bottom-up manner. (b) Once the embedding layer is trained, the embedding layer employs a top-down strategy during the inference process, independently formulating decisions without assistance from any calibration model. (c) Through different training and inf… view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: The three-step bottom-up training process of SenDaL. Freeze notation in unified model training denotes a layer in which weight does not change during backpropagation. layer is adequately trained to make independent decisions. Once the embedding layer has received suffi…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 43 canonical work pages

  1. [1]

    Internet of things-iot: Definition, char- acteristics, architecture, enabling technologies, application & future challenges,

    K. K. Patel and S. M. Patel, “Internet of things-iot: Definition, char- acteristics, architecture, enabling technologies, application & future challenges,” 2016

    work page 2016

  2. [2]

    A survey on stan- dards for interoperability and security in the internet of things,

    E. Lee, Y .-D. Seo, S.-R. Oh, and Y .-G. Kim, “A survey on stan- dards for interoperability and security in the internet of things,” IEEE Communications Surveys & Tutorials, vol. 23, no. 2, pp. 1020–1047, 2021

    work page 2021

  3. [3]

    Robot learning in homes: Improving generalization and reducing dataset bias,

    A. Gupta, A. Murali, D. Gandhi, and L. Pinto, “Robot learning in homes: Improving generalization and reducing dataset bias,” 2018. [Online]. Available: https://arxiv.org/abs/1807.07049

    arXiv 2018

  4. [4]

    Scaling data-driven robotics with reward sketching and batch reinforcement learning,

    S. Cabi, S. G. Colmenarejo, A. Novikov, K. Konyushkova, S. Reed, R. Jeong, K. Zolna, Y . Aytar, D. Budden, M. Vecerik, O. Sushkov, D. Barker, J. Scholz, M. Denil, N. de Freitas, and Z. Wang, “Scaling data-driven robotics with reward sketching and batch reinforcement learning,” 2019. [Online]. Available: https://arxiv.org/abs/1909.12200

    arXiv 2019

  5. [5]

    A survey on learning-based robotic grasping,

    K. Kleeberger, R. Bormann, W. Kraus, and M. F. Huber, “A survey on learning-based robotic grasping,” Current Robotics Reports, vol. 1, no. 4, pp. 239–249, Dec 2020. [Online]. Available: https://doi.org/10.1007/s43154-020-00021-6

    work page doi:10.1007/s43154-020-00021-6 2020

  6. [6]

    A data-driven review of soft robotics,

    B. Jumet, M. D. Bell, V . Sanchez, and D. J. Preston, “A data-driven review of soft robotics,” Advanced Intelligent Systems, vol. 4, no. 4, p. 2100163, 2022

    work page 2022

  7. [7]

    Soft sensor development and applications based on lstm in deep neural networks,

    W. Ke, D. Huang, F. Yang, and Y . Jiang, “Soft sensor development and applications based on lstm in deep neural networks,” 11 2017, pp. 1–6

    work page 2017

  8. [8]

    Soft sensor validation for monitor- ing and resilient control of sequential subway indoor air quality through memory-gated recurrent neural networks-based autoencoders,

    J. Loy-Benitez, S. Heo, and C. Yoo, “Soft sensor validation for monitor- ing and resilient control of sequential subway indoor air quality through memory-gated recurrent neural networks-based autoencoders,” Control Engineering Practice, vol. 97, p. 104330, 2020

    work page 2020

Show all 52 references
  1. [9]

    Soft sensor transferability: A survey,

    F. Curreri, L. Patan ´e, and M. Xibilia, “Soft sensor transferability: A survey,” Applied Sciences, vol. 11, p. 7710, 08 2021

    2021

  2. [10]

    A survey on deep learning for data-driven soft sensors,

    Q. Sun and Z. Ge, “A survey on deep learning for data-driven soft sensors,” IEEE Transactions on Industrial Informatics, vol. 17, no. 9, pp. 5853–5866, 2021. IEEE INTERNET OF THINGS JOURNAL, VOL. 11, NO. 11, 1 JUNE 2024 11

    2021

  3. [11]

    A tinyml soft-sensor approach for low-cost detection and monitoring of vehicular emissions,

    P. Andrade, I. Silva, M. Silva, T. Flores, J. Cassiano, and D. G. Costa, “A tinyml soft-sensor approach for low-cost detection and monitoring of vehicular emissions,” Sensors, vol. 22, no. 10, 2022. [Online]. Available: https://www.mdpi.com/1424-8220/22/10/3838

    2022

  4. [12]

    Long Short-Term Memory,

    S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, 11 1997

    1997

  5. [13]

    On the properties of neural machine translation: Encoder-decoder approaches,

    K. Cho, B. van Merrienboer, D. Bahdanau, and Y . Bengio, “On the properties of neural machine translation: Encoder-decoder approaches,”

  6. [14]

    Phased lstm: Accelerating recurrent network training for long or event-based sequences,

    D. Neil, M. Pfeiffer, and S.-C. Liu, “Phased lstm: Accelerating recurrent network training for long or event-based sequences,” 2016

    2016

  7. [15]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention is all you need,” 2017. [Online]. Available: https://arxiv.org/abs/1706.03762

    2017 arXiv

  8. [16]

    Reformer: The efficient transformer,

    N. Kitaev, Ł. Kaiser, and A. Levskaya, “Reformer: The efficient transformer,” 2020. [Online]. Available: https://arxiv.org/abs/2001.04451

    2020 arXiv

  9. [17]

    Rethinking attention with performers,

    K. Choromanski, V . Likhosherstov, D. Dohan, X. Song, A. Gane, T. Sarlos, P. Hawkins, J. Davis, A. Mohiuddin, L. Kaiser, D. Belanger, L. Colwell, and A. Weller, “Rethinking attention with performers,”

  10. [18]

    Smart home systems based on internet of things,

    M. Domb, “Smart home systems based on internet of things,” in Internet of Things (IoT) for Automated and Smart Applications, Y . Ismail, Ed. Rijeka: IntechOpen, 2019, ch. 3. [Online]. Available: https://doi.org/10.5772/intechopen.84894

    2019 doi

  11. [19]

    Relsen: An optimization-based framework for simultaneously sensor reliability monitoring and data cleaning,

    C. Feng, X. Liang, D. Schneegass, and P. Tian, “Relsen: An optimization-based framework for simultaneously sensor reliability monitoring and data cleaning,” 2020. [Online]. Available: https: //arxiv.org/abs/2004.08762

    2020 arXiv

  12. [20]

    C- lstm: Enabling efficient lstm using structured compression techniques on fpgas,

    S. Wang, Z. Li, C. Ding, B. Yuan, Y . Wang, Q. Qiu, and Y . Liang, “C- lstm: Enabling efficient lstm using structured compression techniques on fpgas,” 2018. [Online]. Available: https://arxiv.org/abs/1803.06305

    2018 arXiv

  13. [21]

    Efficient LSTM training with eligibility traces,

    M. Hoyer, S. Eivazi, and S. Otte, “Efficient LSTM training with eligibility traces,” in Lecture Notes in Computer Science. Springer Nature Switzerland, 2022, pp. 334–346

    2022

  14. [22]

    Energy-efficient lstm inference accelerator for real-time causal prediction,

    Z. Chen, H. T. Blair, and J. Cong, “Energy-efficient lstm inference accelerator for real-time causal prediction,” ACM Trans. Des. Autom. Electron. Syst., vol. 27, no. 5, jun 2022. [Online]. Available: https://doi.org/10.1145/3495006

    2022 doi

  15. [23]

    Informer: Beyond efficient transformer for long sequence time-series forecasting,

    H. Zhou, S. Zhang, J. Peng, S. Zhang, J. Li, H. Xiong, and W. Zhang, “Informer: Beyond efficient transformer for long sequence time-series forecasting,” 2020. [Online]. Available: https://arxiv.org/abs/2012.07436

    2020 arXiv

  16. [24]

    Pyraformer: Low-complexity pyramidal attention for long-range time series modeling and forecasting,

    S. Liu, H. Yu, C. Liao, J. Li, W. Lin, A. X. Liu, and S. Dust- dar, “Pyraformer: Low-complexity pyramidal attention for long-range time series modeling and forecasting,” in International Conference on Learning Representations, 2021

    2021

  17. [25]

    Ecoformer: Energy- saving attention with linear complexity,

    J. Liu, Z. Pan, H. He, J. Cai, and B. Zhuang, “Ecoformer: Energy- saving attention with linear complexity,” 2022. [Online]. Available: https://arxiv.org/abs/2209.09004

    2022 arXiv

  18. [26]

    Transformer hawkes process,

    S. Zuo, H. Jiang, Z. Li, T. Zhao, and H. Zha, “Transformer hawkes process,” 2020. [Online]. Available: https://arxiv.org/abs/2002.09291

    2020 arXiv

  19. [27]

    Ensembles for time series forecasting,

    M. Oliveira and L. Torgo, “Ensembles for time series forecasting,” in Proceedings of the Sixth Asian Conference on Machine Learning, ser. Proceedings of Machine Learning Research, D. Phung and H. Li, Eds., vol. 39. Nha Trang City, Vietnam: PMLR, 26–28 Nov 2015, pp. 360–370. [O...

    2015

  20. [28]

    Time series forecasting using ensemble learning methods for emergency prevention in hydroelectric power plants with dam,

    S. F. Stefenon, M. H. D. M. Ribeiro, A. Nied, K.-C. Yow, V . C. Mariani, L. dos Santos Coelho, and L. O. Seman, “Time series forecasting using ensemble learning methods for emergency prevention in hydroelectric power plants with dam,” Electric Power Systems Research, vol. 202,...

    2022

  21. [29]

    A study on ensemble learning for time series forecasting and the need for meta- learning,

    J. Gastinger, S. Nicolas, D. Stepi ´c, M. Schmidt, and A. Sch¨ulke, “A study on ensemble learning for time series forecasting and the need for meta- learning,” 2021. [Online]. Available: https://arxiv.org/abs/2104.11475

    2021 arXiv

  22. [30]

    Ensemble learning- based modeling and short-term forecasting algorithm for time series with small sample,

    Y . Zhang, G. Ren, X. Liu, G. Gao, and M. Zhu, “Ensemble learning- based modeling and short-term forecasting algorithm for time series with small sample,” Engineering Reports, vol. 4, no. 5, p. e12486, 2022

    2022

  23. [31]

    Feature-weighted stacking for nonsea- sonal time series forecasts: A case study of the COVID-19 epidemic curves,

    P. Cawood and T. L. van Zyl, “Feature-weighted stacking for nonsea- sonal time series forecasts: A case study of the COVID-19 epidemic curves,” in 2021 8th International Conference on Soft Computing & Machine Intelligence (ISCMI). IEEE, nov 2021

    2021

  24. [32]

    A hybrid gpu-fpga based design methodology for enhancing machine learning applications performance,

    X. Liu, H.-A. Ounifi, A. Gherbi, W. Li, and M. Cheriet, “A hybrid gpu-fpga based design methodology for enhancing machine learning applications performance,” Journal of Ambient Intelligence and Humanized Computing, vol. 11, no. 6, pp. 2309–2323, Jun 2020. [Online]. Available: ...

    2020 doi

  25. [33]

    Tinyml- enabled edge implementation of transfer learning framework for domain generalization in machine fault diagnosis,

    S. Asutkar, C. Chalke, K. Shivgan, and S. Tallur, “Tinyml- enabled edge implementation of transfer learning framework for domain generalization in machine fault diagnosis,” Expert Systems with Applications, vol. 213, p. 119016, 2023. [Online]. Available: https://www.sciencedir...

    2023

  26. [34]

    Tensorflow lite micro: Embedded machine learning on tinyml systems,

    R. David, J. Duke, A. Jain, V . J. Reddi, N. Jeffries, J. Li, N. Kreeger, I. Nappier, M. Natraj, S. Regev, R. Rhodes, T. Wang, and P. Warden, “Tensorflow lite micro: Embedded machine learning on tinyml systems,” 2021

    2021

  27. [35]

    An opencl(tm) deep learning accelerator on arria 10,

    U. Aydonat, S. O’Connell, D. Capalija, A. C. Ling, and G. R. Chiu, “An opencl(tm) deep learning accelerator on arria 10,” 2017

    2017

  28. [36]

    Sensorformer: Efficient many- to-many sensor calibration with learnable input subsampling,

    Y . Cheng, O. Saukh, and L. Thiele, “Sensorformer: Efficient many- to-many sensor calibration with learnable input subsampling,” IEEE Internet of Things Journal, vol. 9, no. 20, pp. 20 577–20 589, 2022

    2022

  29. [37]

    Smart plants: Low-cost solution for monitoring indoor environments,

    A. Zuniga, N. H. Motlagh, H. Flores, and P. Nurmi, “Smart plants: Low-cost solution for monitoring indoor environments,” IEEE Internet of Things Journal, vol. 9, no. 22, pp. 23 252–23 259, 2022

    2022

  30. [38]

    Predict- ing low-cost gas sensor readings from transients using long short-term memory neural networks,

    J. ˇCuli´c Gambiroˇza, T. Masteli´c, T. Kovaˇcevi´c, and M. ˇCagalj, “Predict- ing low-cost gas sensor readings from transients using long short-term memory neural networks,” IEEE Internet of Things Journal, vol. 7, no. 9, pp. 8451–8461, 2020

    2020

  31. [39]

    Low-cost co sensor calibration using one dimensional convolutional neural network,

    S. Ali, F. Alam, K. Arif, and J. Potgieter, “Low-cost co sensor calibration using one dimensional convolutional neural network,” Sensors, vol. 23, 01 2023

    2023

  32. [40]

    Two step calibration method for ozone low-cost sensor: Field experiences with the urbansense dcus,

    J. S ´a, H. Chojer, P. Branco, M. Alvim-Ferraz, F. Martins, and S. Sousa, “Two step calibration method for ozone low-cost sensor: Field experiences with the urbansense dcus,” Journal of Environmental Management, vol. 328, p. 116910, 2023. [Online]. Available: https://www.scien...

    2023

  33. [41]

    Is smarter better? a comparison of adaptive, and simple moving average trading strategies,

    C. A. Ellis and S. A. Parbery, “Is smarter better? a comparison of adaptive, and simple moving average trading strategies,” Research in International Business and Finance, vol. 19, no. 3, pp. 399–411, 2005

    2005

  34. [42]

    Postwar u.s. business cycles: An empirical investigation,

    R. J. Hodrick and E. C. Prescott, “Postwar u.s. business cycles: An empirical investigation,” Journal of Money, Credit and Banking, vol. 29, no. 1, pp. 1–16, 1997. [Online]. Available: http://www.jstor. org/stable/2953682

    1997

  35. [43]

    Rectifier nonlinearities improve neural network acoustic models,

    A. L. Maas, A. Y . Hannun, and A. Y . Ng, “Rectifier nonlinearities improve neural network acoustic models,” in in ICML Workshop on Deep Learning for Audio, Speech and Language Processing, 2013

    2013

  36. [44]

    Low-cost outdoor air quality monitoring and sensor calibration: A survey and critical analysis,

    F. Concas, J. Mineraud, E. Lagerspetz, S. Varjonen, X. Liu, K. Puo- lam¨aki, P. Nurmi, and S. Tarkoma, “Low-cost outdoor air quality monitoring and sensor calibration: A survey and critical analysis,” ACM Trans. Sen. Netw., vol. 17, no. 2, may 2021

    2021

  37. [45]

    Field evaluation of low-cost particulate matter sensors in high- and low-concentration environments,

    T. Zheng, M. H. Bergin, K. K. Johnson, S. N. Tripathi, S. Shirodkar, M. S. Landis, R. Sutaria, and D. E. Carlson, “Field evaluation of low-cost particulate matter sensors in high- and low-concentration environments,” Atmospheric Measurement Techniques, vol. 11, no. 8, pp. 4823...

    2018

  38. [46]

    Long-term field comparison of multiple low-cost particulate matter sensors in an outdoor urban environment,

    F. M. J. Bulot, S. J. Johnston, P. J. Basford, N. H. C. Easton, M. Apetroaie-Cristea, G. L. Foster, A. K. R. Morris, S. J. Cox, and M. Loxham, “Long-term field comparison of multiple low-cost particulate matter sensors in an outdoor urban environment,” Scientific Reports, vol....

    2019 doi

  39. [47]

    A review of low-cost particulate matter sensors from the developers’ perspectives,

    B. Alfano, L. Barretta, A. Del Giudice, S. De Vito, G. Di Francia, E. Esposito, F. Formisano, E. Massera, M. L. Miglietta, and T. Polichetti, “A review of low-cost particulate matter sensors from the developers’ perspectives,” Sensors, vol. 20, no. 23, 2020. [Online]. Availabl...

    2020

  40. [48]

    D. A. Freedman, Statistical Models: Theory and Practice, 2nd ed. Cambridge University Press, 2009

    2009

  41. [49]

    Machine learning for fore- casting mid-price movements using limit order book data,

    P. Nousi, A. Tsantekidis, N. Passalis, A. Ntakaris, J. Kanniainen, A. Tefas, M. Gabbouj, and A. Iosifidis, “Machine learning for fore- casting mid-price movements using limit order book data,” Ieee Access, vol. 7, pp. 64 722–64 736, 2019

    2019

  42. [50]

    Rmse is not enough: Guidelines to robust data-model comparisons for magnetospheric physics,

    M. W. Liemohn, A. D. Shane, A. R. Azari, A. K. Petersen, B. M. Swiger, and A. Mukhopadhyay, “Rmse is not enough: Guidelines to robust data-model comparisons for magnetospheric physics,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 218, p. 105624, 2021. [Online]. ...

    2021

  43. [2014]

    Available: https://arxiv.org/abs/1409.1259

    [Online]. Available: https://arxiv.org/abs/1409.1259

    arXiv

  44. [2021]

    Available: https://arxiv.org/abs/2009.14794

    [Online]. Available: https://arxiv.org/abs/2009.14794

    2009 arXiv

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.