Pith. sign in

REVIEW 4 major objections 5 minor 66 references

In-field Calibration of Low-Cost Sensors through XGBoost $\&$ Aggregate Sensor Data

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A spatial XGBoost model trained on one city's low-cost PM2.5 sensors can calibrate new deployments with only a short fine-tuning step.

desk verdict Target leakage likely makes the reported RMSEs circular, and the paper's own zero-shot result contradicts its cross-location generalization claim. read the letter →

arxiv 2506.15840 v1 pith:7J4ENZAB submitted 2025-06-18 cs.LG

Kevin Yin , Julia Gersey , Pei Zhang This is my paper · ORCID
classification cs.LG
keywords XGBoostsensorcalibrationPM2.5low-costairqualitysensorsspatialgeneralizationfine-tuningensemblelearningSensEURCity
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

This paper tries to show that a single XGBoost calibration model, trained on readings from many low-cost particulate matter sensors in one city, can be reused across new sensor deployments and even new cities with minimal additional tuning. The motivation is that low-cost air quality sensors drift and need calibration against expensive reference stations, which limits how widely they can be deployed. The authors claim their model, using location, temperature, humidity, and reference readings from neighboring sensors as inputs, reaches an RMSE of 6.41 µg/m³ for new sensors added within the learned area after only about 100 boosting rounds of fine-tuning, and 6.52 µg/m³ for a completely new city. If true, this would make large-scale air quality monitoring cheaper by letting a calibration model transfer from one deployment to the next.

What carries the argument

The central object is an XGBoost gradient-boosted tree model (gbtree booster) trained on eight input features per sensor reading: the Alphasense OPCN3PM25 particle counter output, the reference PM2.5 measurement, longitude, latitude, internal and external temperature, and internal and external humidity. The geographic coordinates are what let the model learn a spatial mapping of calibration values, while the environmental variables capture known humidity and temperature dependencies of particle sensors. The mechanism that carries the argument is fine-tuning: starting from the Antwerp-trained model, the authors retrain for only 100 boosting rounds on data from a new deployment or new city, which they show is enough to substantially lower RMSE compared to using the original model unmodified.

What would settle it

Run the trained model on a new sensor deployment while withholding or estimating the reference PM2.5 input; if the calibration error degrades sharply compared to the reported RMSE of 6.41, the claim that the model reduces dependence on reference infrastructure is not supported.

Watch

Extended reading notes

Core claim

The paper's central claim is that an XGBoost regression model can act as a spatially aware calibration function for low-cost PM2.5 sensors, learning how calibration adjustments depend on geographic position and local environmental conditions. Trained on data from 34 sensors in Antwerp, the model predicts calibration values for held-out test readings with RMSE 5.248 µg/m³. When applied to new sensor deployments within the same city, minimal fine-tuning for 100 boosting rounds improves performance to RMSE 6.41 µg/m³, comparable to the original validation error. When transferred to entirely different cities such as Oslo and Zagreb, the model initially fails with RMSE around 250 µg/m³, but fine-tuning it for about 100 rounds on a small amount of local data brings RMSE down to 6.52 µg/m³. The authors conclude that the model learns spatial relationships during initial training and can be quickly adapted to new locations, reducing the need for extensive recalibration.

Load-bearing premise

The model requires the high-accuracy reference PM2.5 measurement as an input feature at every sensor location at inference time, which is exactly the expensive infrastructure the method aims to avoid.

Editorial extensions

If this is right

  • A calibration model trained once on a dense sensor network can be reused for new sensors in the same area with only about 100 rounds of fine-tuning, reaching RMSE 6.41 µg/m³.
  • A model trained in one city can be adapted to a new city by fine-tuning on a small amount of local data, reaching RMSE 6.52 µg/m³ on Oslo and Zagreb.
  • Increasing the number of deployment locations in the training set improves the model's ability to generalize to unseen locations.
  • The method reduces the data collection and tuning burden for calibrating new low-cost sensor deployments, making it easier to expand air quality monitoring networks.

Reading between the lines

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

  • The paper does not test a scenario where the reference PM2.5 value is absent at inference time, so the practical claim of reducing dependence on expensive reference stations remains unverified; a natural extension would be to train a model variant that omits Ref.PM2.5 or imputes it from neighboring stations.
  • Since the model does not rely on sensor type, the same spatial calibration approach could in principle be applied to gaseous pollutant sensors, but this generalization is speculative until tested.
  • The fine-tuning procedure described uses only the number of boosting rounds; an ablation varying the amount of local data and the number of tuned hyperparameters would clarify how little data is actually needed for transfer.
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.

Referee Report

4 major / 5 minor

Summary. The paper proposes an XGBoost-based spatial calibration model for low-cost PM2.5 sensors, trained on the SenEURCity dataset from Antwerp, Oslo, and Zagreb. The model is intended to calibrate a sensor using readings from neighboring deployments plus environmental variables. The authors report an RMSE of 5.248 on the Antwerp test set, RMSE of 250.69 when applied directly to Oslo without fine-tuning, and RMSE values of 6.52 and 6.41 after fine-tuning for new cities and for new sensors within a known network, respectively. The central claim is that the model reduces dependence on expensive reference infrastructure and generalizes across locations.

Significance. If the reported results were obtained without target leakage, the paper would offer a useful step toward scalable in-field calibration of low-cost sensor networks using public data and open code. The use of the public SenEURCity dataset, the explicit release of code, and the focus on spatial generalization are positive aspects. However, the significance is substantially weakened by the ambiguous definition of the target variable and by the inclusion of the reference PM2.5 measurement as an input feature; these issues directly affect whether the reported RMSE values measure genuine calibration rather than trivial reproduction of the target.

major comments (4)
  1. [Section 3.2] The paper lists Ref.PM2.5 (the reference PM2.5 measurement) as one of the eight input features, but the target variable is never explicitly defined anywhere in the manuscript. In Section 4, the reported RMSE is described as measuring error against 'ground truth calibration values,' yet it is not stated whether the target is the reference PM2.5 value itself, a correction term such as Ref.PM2.5 minus the low-cost reading, or something else. If the target is Ref.PM2.5, the model is being asked to predict a feature it already receives, and a tree ensemble can trivially track that feature. If the target is a correction derived from Ref.PM2.5, then the model can compute the answer almost directly from that single input. The authors must define the target precisely and must report an ablation that removes Ref.PM2.5 from the feature set; without such an ablation, the reported RMSE values cannot be interpreted.
  2. [Section 4] The paper's central claim of reducing dependence on reference infrastructure is contradicted by its own no-fine-tuning result on Oslo, where RMSE is 250.69. The positive generalization results (RMSE 6.52 and 6.41) are obtained only after fine-tuning on reference data from the new location or new sensors. The inference-time protocol is not specified: when a newly deployed sensor lacks a co-located reference station, either the model cannot be applied because Ref.PM2.5 is missing from the input vector, or it must be applied under a feature distribution different from training. The paper needs to explain how the model is used in the intended deployment scenario where no reference measurement is available, and it needs to report performance without fine-tuning in a way that is consistent with the stated goal.
  3. [Section 4] The evaluation methodology is insufficient to support the cross-location generalization claim. The experiments use a single train/validation/test split from a single city (Antwerp) for the base model, with no error bars, no repeated runs, and no comparison to standard baselines such as the raw low-cost sensor readings, simple linear regression, monosensor XGBoost, or the spatial XGBoost mapping model from prior work [51]. The statement that 'the RMSE is a summation of individual sensor errors' is also unclear because RMSE is typically defined as the square root of the mean squared error across all predictions, not a sum. Without baselines and a clear definition of the evaluation metric, the reported RMSE values do not establish that the proposed model improves over existing calibration or spatial mapping methods.
  4. [Section 3.3] The description of hyperparameter tuning is incomplete and potentially misleading. The text says the model uses XGBoost's 'built in hyperparameter turning,' then describes a grid search, but only learning rate (0.16) and n_estimators (500) are mentioned as non-default, with default values used for all other hyperparameters. It is not stated whether the reported test RMSE was selected based on the validation set, how many configurations were tried, whether results were stable across random seeds, or how the 480,000 data points from 34 sensors were distributed across train, validation, and test splits. This lack of detail makes it impossible to assess the risk of overfitting or selection bias in the reported performance.
minor comments (5)
  1. [Section 1] There are several typographical and grammatical errors, including 'it's reduced sizing' (should be 'its'), 'course particulates' (should be 'coarse particulates'), and 'golden standard' (should be 'gold standard'). These do not affect the technical content but should be corrected.
  2. [Section 3.2] The sentence 'Of these, the highest RMSE was found to result from forward and backward filling' is likely the opposite of what is intended, since the next clause says this method 'better captur[es] more realistic gradual change.' The text should state which preprocessing gave the lowest RMSE, with the corresponding number.
  3. [Section 3.2] The abstract and introduction say the model 'consolidates data from neighboring sensors,' but Section 3.2 states that other sensor types were excluded to avoid reliance on co-deployed sensors. The manuscript should clarify what 'neighboring sensor data' means here: the input vector appears to contain only location coordinates, temperature, humidity, and the target sensor's own PM2.5 reading, not readings from other low-cost sensors.
  4. [Section 4] Figure 2 and Figure 3 show training and validation RMSE during fine-tuning but lack axis labels, legends, and a description of the fine-tuning setup (e.g., whether the original model weights were frozen except for the last trees, or the entire model was retrained). Adding these details would improve reproducibility.
  5. [References] Several references are incomplete or formatted inconsistently; for example, [55] does not include the full author list or title of the SensEURCity data descriptor, and the reference for the global burden of disease study [32] lacks a full journal name. The reference list should be checked against the publisher's guidelines.

Circularity Check

2 steps flagged · score 7.0 of 10

The reported calibration RMSE is uninformative because Ref.PM2.5 is both an input feature and, through the undefined 'calibration value' target, the quantity being predicted; the cross-location results are additionally obtained after fine-tuning on target reference data.

  1. self definitional [Section 3.2 (Data Preprocessing) and Section 4 (Results); the regression target is never defined.]
    "In the end, 8 input variables are taken from each low-cost sensor location: OPCN3PM25(Alphasense PM2.5 counter), Ref.PM2.5(reference PM2.5 measurement), longitude, latitude, SHT31TI(internal temperature), SHT31TE(external temperature), SHT31HI(internal humidity), and SHT31HE(external humidity). Reference reading is taken to provide a baseline for training."

    The paper's goal is to 'predict calibration adjustments across the sensor network' and it reports RMSE against 'ground truth calibration values', but it never defines the regression target. If the target is Ref.PM2.5, the model receives the answer as an input feature and a tree ensemble can copy it. If the target is a correction such as Ref.PM2.5 minus the low-cost OPCN3PM25 reading, both terms of that difference are also input features, so the model can compute the target almost directly from a single feature. In either reading, the reported RMSE of 5.248 is forced by construction and provides no evidence of in-field calibration without a co-located reference. No ablation dropping Ref.PM2.5 is reported, and no inference protocol is given for deployments where Ref.PM2.5 is absent.

  2. fitted input called prediction [Section 4 (Results), fine-tuning experiments before Figures 2 and 3.]
    "We find that with minimal fine turning for only 100 boosting rounds on the tree model, calibration model performance on the collocations can be improved to a lowest RMSE of 6.52, with additional fine-tuning resulting in marginally improved RMSE as seen in figure 2."

    This 'generalization to new sensor network deployments' result is obtained after fine-tuning on reference/co-located data at the new location ('collocations'). Likewise, the within-network addition of two sensors reaches RMSE 6.41 only after 'minimal fine tuning for about 100 boosting rounds.' The headline claim that the model 'requires minimal fine tuning to new locations outside the learned area' therefore depends on the availability of high-accuracy reference measurements at the target location—the same expensive infrastructure the method claims to reduce dependence on. The reported RMSEs are post-fit numbers, not predictions made before seeing target reference data.

full rationale

The central calibration result is compromised by definitional target leakage. The paper lists Ref.PM2.5, the reference PM2.5 measurement, among the eight input features in Section 3.2, and it evaluates against 'ground truth calibration values' without ever defining the target variable. If the target is the reference PM2.5 value, the model is asked to predict a feature it already receives; if the target is a calibration adjustment derived from the reference, the reference feature still nearly determines the target. Either way, the reported RMSE of 5.248 does not measure the model's ability to calibrate low-cost sensors without reference infrastructure. The deployment story is also incomplete: no inference-time protocol is described for sensors lacking a co-located reference station, so a deployed model either cannot be applied when Ref.PM2.5 is missing or operates under a different feature distribution. The generalization claims are further weakened because the cross-city and new-sensor results are produced after fine-tuning on the target location's reference or collocation data. There is no significant self-citation-based circularity; the problem is that the paper's own feature set makes the prediction target redundant by construction. Because the main accuracy result reduces to a feature already present in the input, the circularity score is 7.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new entities. Its central claim rests on the availability and accuracy of reference measurements, on the sufficiency of the chosen spatial and environmental features, and on the validity of treating temporally correlated readings as independent samples. No machine-checked proofs, shipped data artifacts, or parameter-free derivations are provided.

free parameters (3)
  • XGBoost learning rate = 0.16
    Tuned by grid search on the validation set (Section 3.3).
  • XGBoost n_estimators = 500
    Tuned by grid search, trained for 1000 rounds (Section 3.3).
  • Default XGBoost hyperparameters = default
    max_depth, subsample, colsample_bytree, min_child_weight left at defaults; still model choices that affect results.
assumptions (4)
  • domain assumption Reference stations provide accurate PM2.5 ground truth
    The model is trained and evaluated against Ref.PM2.5 values from high-accuracy stations (SensEURCity).
  • domain assumption Latitude and longitude are sufficient spatial features for calibration transfer
    The model relies on these coordinates to learn spatial relationships (Section 3.2).
  • domain assumption Temperature and humidity capture environmental sensitivity
    The authors include SHT31 T/H readings because environmental variables influence sensor output (Section 3.2).
  • domain assumption Readings in the dataset are independent for train/test splits
    The data is split along time without accounting for temporal autocorrelation within each sensor deployment (Section 3.2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of In-field Calibration of Low-Cost Sensors through XGBoost $\&$ Aggregate Sensor Data." pith.science (2026). https://pith.science/paper/7J4ENZAB

@misc{pith2026250615840,
  author       = {Pith},
  title        = {Pith review of: In-field Calibration of Low-Cost Sensors through XGBoost $\&$ Aggregate Sensor Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7J4ENZAB}},
  note         = {Machine review of arXiv:2506.15840}
}
read the original abstract

Effective large-scale air quality monitoring necessitates distributed sensing due to the pervasive and harmful nature of particulate matter (PM), particularly in urban environments. However, precision comes at a cost: highly accurate sensors are expensive, limiting the spatial deployments and thus their coverage. As a result, low-cost sensors have become popular, though they are prone to drift caused by environmental sensitivity and manufacturing variability. This paper presents a model for in-field sensor calibration using XGBoost ensemble learning to consolidate data from neighboring sensors. This approach reduces dependence on the presumed accuracy of individual sensors and improves generalization across different locations.

Figures

Figures reproduced from arXiv: 2506.15840 by the authors.

Figure 1
Figure 1. Train and validation RMSE during training [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Train and validation RMSE during fine tuning for [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 59 canonical work pages

  1. [51]

    Yingqiang Song, Changjian Zhang, Xin Jin, Xiaoyu Zhao, Wei Huang, Xiaoshuang Sun, Zhongkang Yang, and Shuhuan Wang. 2023. Spatial prediction of PM2.5 concentration using hyper-parameter optimization XGBoost model in China. 32 (2023), 103272. doi:10.1016/j.eti.2023.103272

    work page arXiv 2023

  2. [1]

    Marie-Laure Aix, Seán Schmitz, and Dominique J. Bicout. 2023. Calibration methodology of low-cost sensors for high-quality monitoring of fine particulate matter. 889 (2023), 164063. doi:10.1016/j.scitotenv.2023.164063

    work page arXiv 2023

  3. [2]

    Anderson, Josef G

    Jonathan O. Anderson, Josef G. Thundiyil, and Andrew Stolbach. 2012. Clearing the Air: A Review of the Effects of Particulate Matter Air Pollution on Human Health. 8, 2 (2012), 166–175. doi:10.1007/s13181-011-0203-1

    work page doi:10.1007/s13181-011-0203-1 2012

  4. [3]

    Matarazzo, Priyanka deSouza, and Carlo Ratti

    Amin Anjomshoaa, Fábio Duarte, Daniël Rennings, Thomas J. Matarazzo, Priyanka deSouza, and Carlo Ratti. 2018. City Scanner: Building and Sched- uling a Mobile Sensing Platform for Smart City Services.IEEE Internet of Things Journal5, 6 (Dec. 2018), 4567–4579. doi:10.1109/JIOT.2018.2839058

    work page arXiv 2018

  5. [4]

    Apte, Kyle P

    Joshua S. Apte, Kyle P. Messier, Shahzad Gani, Michael Brauer, Thomas W. Kirch- stetter, Melissa M. Lunden, Julian D. Marshall, Christopher J. Portier, Roel C.H. Vermeulen, and Steven P. Hamburg. 2017. High-Resolution Air Pollution Map- ping with Google Street View Cars: Exploiting Big Data. 51, 12 (2017), 6999–7008. doi:10.1021/acs.est.7b00891

    work page doi:10.1021/acs.est.7b00891 2017

  6. [5]

    Claudio Badii, Stefano Bilotta, Daniele Cenni, Angelo Difino, Paolo Nesi, Irene Paoli, and Michela Paolucci. 2020. Real-Time Automatic Air Pollution Services from IOT Data Network. In2020 IEEE Symposium on Computers and Communi- cations (ISCC). 1–6. doi:10.1109/ISCC50000.2020.9219580 ISSN: 2642-7389

    arXiv 2020

  7. [6]

    Marzena Banach, Tomasz Talaśka, Jakub Dalecki, and Rafał Długosz

    work page

  8. [7]

    Virendra Barot and Viral Kapadia. 2020. Air quality monitoring systems using IoT: a review. In2020 international conference on computational performance evaluation (ComPE). IEEE, 226–231

    work page 2020

Show all 66 references
  1. [8]

    Kinney, Matthew S

    Jing Cai, Beizhan Yan, James Ross, Danian Zhang, Patrick L. Kinney, Matthew S. Perzanowski, KyungHwa Jung, Rachel Miller, and Steven N. Chillrud. 2014. Validation of MicroAeth®as a Black Carbon Monitor for Fixed-Site Measurement and Optimization for Personal Exposure Character...

    2014 doi

  2. [9]

    Sreeni Chadalavada, Oliver Faust, Massimo Salvi, Silvia Seoni, Nawin Raj, U Raghavendra, Anjan Gudigar, Prabal Datta Barua, Filippo Molinari, and Rajendra Acharya. 2024. Application of artificial intelligence in air pollution monitoring and forecasting: A systematic review.Env...

    2024

  3. [10]

    Xinlei Chen, Susu Xu, Haohao Fu, Carlee Joe-Wong, Lin Zhang, Hae Young Noh, and Pei Zhang. 2019. ASC: actuation system for city-wide crowdsensing with ride-sharing vehicular platform. InProceedings of the Fourth Workshop on International Science of Smart City Operations and Pl...

    2019

  4. [11]

    Xinlei Chen, Xiangxiang Xu, Xinyu Liu, Shijia Pan, Jiayou He, Hae Young Noh, Lin Zhang, and Pei Zhang. 2018. Pga: Physics guided and adaptive approach for mobile fine-grained air pollution estimation. InProceedings of the 2018 ACM International Joint Conference and 2018 Intern...

    2018

  5. [12]

    Yun Cheng, Xiaoxi He, Zimu Zhou, and Lothar Thiele. 2020. MapTransfer : Urban Air Quality Map Generation for Downscaled Sensor Deployments. In 2020 IEEE/ACM Fifth International Conference on Internet-of-Things Design and Implementation (IoTDI). 14–26. doi:10.1109/IoTDI49375.2020.00010

    2020

  6. [13]

    Yun Cheng, Xiucheng Li, Zhijun Li, Shouxu Jiang, Yilong Li, Ji Jia, and Xiaofan Jiang. 2014. AirCloud: a cloud-based air-quality monitoring system for everyone. InProceedings of the 12th ACM Conference on Embedded Network Sensor Systems. ACM, Memphis Tennessee, 251–265. doi:10...

    2014

  7. [14]

    Hone-Jay Chu, Muhammad Zeeshan Ali, and Yu-Chen He. 2020. Spatial calibra- tion and PM2.5 mapping of low-cost air quality sensors.Scientific Reports10, 1 (Dec. 2020), 22079. doi:10.1038/s41598-020-79064-w

    2020 doi

  8. [15]

    Gentner, and Kirsten Koehler

    Abhirup Datta, Arkajyoti Saha, Misti Levy Zamora, Colby Buehler, Lei Hao, Fulizi Xiong, Drew R. Gentner, and Kirsten Koehler. 2020. Statistical field calibration of a low-cost PM2.5 monitoring network in Baltimore. 242 (2020), 117761. doi:10. 1016/j.atmosenv.2020.117761

    2020

  9. [16]

    Aoki, Neil Kumar, Alan Mainwaring, Chris Myers, Wesley Willett, and Allison Woodruff

    Prabal Dutta, Paul M. Aoki, Neil Kumar, Alan Mainwaring, Chris Myers, Wesley Willett, and Allison Woodruff. 2009. Common Sense: participatory urban sensing using a network of handheld air quality monitors. InProceedings of the 7th ACM In-field Calibration of Low-Cost Sensors t...

    2009

  10. [17]

    Environmental Protection Agency. 2022. Health and Environmental Ef- fects of Particulate Matter (PM). https://www.epa.gov/pm-pollution/health- andenvironmental-effects-particulate-matter-pm. Accessed: 2025-03-29

    2022

  11. [18]

    Environmental Protection Agency. 2022. Particulate Matter (PM) Basics. https: //www.epa.gov/pm-pollution/particulate-matter-pm-basics. Accessed: 2025-03- 29

    2022

  12. [19]

    Yi Gao, Wei Dong, Kai Guo, Xue Liu, Yuan Chen, Xiaojin Liu, Jiajun Bu, and Chun Chen. 2016. Mosaic: A low-cost mobile sensing system for urban air quality monitoring. InIEEE INFOCOM 2016 - The 35th Annual IEEE International Conference on Computer Communications. 1–9. doi:10.11...

    2016 doi

  13. [20]

    Julia Gersey, Jatin Aggarwal, Jiale Zhang, Jesse Codling, and Pei Zhang. 2025. Sniffing Out the City-Vehicular Multimodal Sensing for Environmental and Infrastructure Analysis. InProceedings of the 23rd ACM Conference on Embedded Networked Sensor Systems. 632–633

    2025

  14. [21]

    Julia Gersey, Rose Allegrette, Joshua Lian, Zawad Munshi, and Aarti Phatke

  15. [22]

    Julia Gersey, Brian Krupp, and Jonathon Fagert. 2023. Pilot Study of Deploying IoT Micro Air Quality Sensors in an Urban Environment: Lessons Learned.J. Comput. Sci. Coll.39, 4 (Oct. 2023), 74–83

    2023

  16. [23]

    A Grantz, J

    D. A Grantz, J. H. B Garner, and D. W Johnson. 2003. Ecological effects of particulate matter. 29, 2 (2003), 213–239. doi:10.1016/S0160-4120(02)00181-2

    2003 doi

  17. [24]

    David Hasenfratz, Olga Saukh, Silvan Sturzenegger, and Lothar Thiele. 2012. Participatory Air Pollution Monitoring Using Smartphones. https://tik-old.ee. ethz.ch/file//b6c2122d089d2ef88348a74ddf2906dc/HSST2012.pdf

    2012

  18. [25]

    Jiayi Huang, Ning Liu, Rui Ma, Xinyu Liu, Yue Wang, and Lin Zhang. 2019. Un- derstanding air pollution patterns in city based on minute-level event detection. InProceedings of the 17th Conference on Embedded Networked Sensor Systems. 448–449

    2019

  19. [26]

    Wan Jiao, Gayle Hagler, Ronald Williams, Robert Sharpe, Ryan Brown, Daniel Garver, Robert Judge, Motria Caudill, Joshua Rickard, Michael Davis, Lewis Wein- stock, Susan Zimmer-Dauphinee, and Ken Buckley. 2016. Community Air Sensor Network (CAIRSENSE) project: evaluation of low...

    2016 doi

  20. [27]

    2021.Taxi-for-all: Incentivized Taxi Actuation System for Balanced Area-wide Service

    Carlee Joe-Wong, Haeyoung Noh, Pei Zhang, and Carnegie Mellon University Mobility21. 2021.Taxi-for-all: Incentivized Taxi Actuation System for Balanced Area-wide Service. Technical Report. https://rosap.ntl.bts.gov/view/dot/59923

    2021

  21. [28]

    Jiwoo Kang and Kanghyeok Choi. 2024. Calibration Methods for Low-Cost Particulate Matter Sensors Considering Seasonal Variability. 24, 10 (2024), 3023. doi:10.3390/s24103023

    2024 doi

  22. [29]

    Brian Krupp, Julia Gersey, Jonathon Fagert, and Tony Mlady. 2023. Towards Fine-Grained Air Quality Sensing in Urban Environments. InProceedings of the 20th ACM Conference on Embedded Networked Sensor Systems (SenSys ’22). Association for Computing Machinery, New York, NY, USA,...

    2023

  23. [30]

    Fung, Julien Mineraud, Samu Varjonen, Matti Siekkinen, Petteri Nurmi, Yutaka Matsumi, Sasu Tarkoma, and Tareq Hussein

    Eemil Lagerspetz, Naser Hossein Motlagh, Martha Arbayani Zaidan, Pak L. Fung, Julien Mineraud, Samu Varjonen, Matti Siekkinen, Petteri Nurmi, Yutaka Matsumi, Sasu Tarkoma, and Tareq Hussein. 2019. MegaSense: Feasibility of Low- Cost Sensors for Pollution Hot-spot Detection. In...

    2019

  24. [31]

    Chieh-Han Lee, Yeuh-Bin Wang, and Hwa-Lung Yu. 2019. An efficient spatiotem- poral data calibration approach for the low-cost PM2.5 sensing network: A case study in Taiwan. 130 (2019), 104838. doi:10.1016/j.envint.2019.05.032

    2019 doi

  25. [32]

    Jos Lelieveld, Andy Haines, Richard Burnett, Cathryn Tonne, Klaus Klingmüller, Thomas Münzel, and Andrea Pozzer. 2023. Air pollution deaths attributable to fossil fuels: observational and modelling study. (2023), e077784. doi:10.1136/bmj- 2023-077784

    2023 doi

  26. [33]

    Shouxu Lin, Yuhang Yao, Pei Zhang, Hae Young Noh, and Carlee Joe-Wong

  27. [34]

    Xinyu Liu, Xinlei Chen, Xiangxiang Xu, Enhan Mai, Hae Young Noh, Pei Zhang, and Lin Zhang. 2017. Delay Effect in Mobile Sensing System for Urban Air Pollution Monitoring. InProceedings of the 15th ACM Conference on Embedded Network Sensor Systems. ACM, Delft Netherlands, 1–2. ...

    2017 doi

  28. [35]

    Xinyu Liu, Xiangxiang Xu, Xinlei Chen, Enhan Mai, Hae Young Noh, Pei Zhang, and Lin Zhang. 2017. Individualized Calibration of Industrial-Grade Gas Sensors in Air Quality Sensing System. InProceedings of the 15th ACM Conference on Embedded Network Sensor Systems. ACM, Delft Ne...

    2017

  29. [36]

    Yuxuan Liu, Haoyang Wang, Fanhang Man, Jingao Xu, Fan Dang, Yunhao Liu, Xiao-Ping Zhang, and Xinlei Chen. 2024. MobiAir: Unleashing Sensor Mobility for City-scale and Fine-grained Air-Quality Monitoring with AirBERT. InProceed- ings of the 22nd Annual International Conference ...

    2024

  30. [37]

    Rui Ma, Ning Liu, Xiangxiang Xu, Yue Wang, Hae Young Noh, Pei Zhang, and Lin Zhang. 2019. A deep autoencoder model for pollution map recovery with mobile sensing networks. InAdjunct Proceedings of the 2019 ACM international joint conference on pervasive and ubiquitous computin...

    2019

  31. [38]

    Rui Ma, Ning Liu, Xiangxiang Xu, Yue Wang, Hae Young Noh, Pei Zhang, and Lin Zhang. 2020. Enhancing the Data Learning With Physical Knowledge in Fine-Grained Air Pollution Inference.IEEE Access8 (2020), 88372–88384. doi:10. 1109/ACCESS.2020.2993610 Conference Name: IEEE Access

    2020

  32. [39]

    Rui Ma, Ning Liu, Xiangxiang Xu, Yue Wang, Hae Young Noh, Pei Zhang, and Lin Zhang. 2020. Fine-Grained Air Pollution Inference with Mobile Sensing Systems: A Weather-Related Deep Autoencoder Model.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.4, 2 (June 2020), 52:1–52:2...

    2020 doi

  33. [40]

    Balz Maag, Zimu Zhou, and Lothar Thiele. 2018. A Survey on Sensor Calibration in Air Pollution Monitoring Deployments. 5, 6 (2018), 4857–4870. doi:10.1109/ JIOT.2018.2853660

    2018

  34. [41]

    Balz Maag, Zimu Zhou, and Lothar Thiele. 2018. W-Air: Enabling Personal Air Pollution Monitoring on Wearables.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.2, 1 (March 2018), 24:1–24:25. doi:10.1145/3191756

    2018 doi

  35. [42]

    Manuel Méndez, Mercedes G Merayo, and Manuel Núñez. 2023. Machine learning algorithms to forecast air quality: a survey.Artificial Intelligence Review56, 9 (2023), 10031–10066

    2023

  36. [43]

    Broday, and Barak Fishbain

    Sharon Moltchanov, Ilan Levy, Yael Etzion, Uri Lerner, David M. Broday, and Barak Fishbain. 2015. On the feasibility of measuring urban air pollution by wireless distributed sensor networks.Science of The Total Environment502 (Jan. 2015), 537–547. doi:10.1016/j.scitotenv.2014.09.059

    2015 doi

  37. [44]

    Adam Moore, Miguel Figliozzi, and Christopher M. Monsere. 2012. Air Quality at Bus Stops: Empirical Analysis of Exposure to Particulate Matter at Bus Stop Shelters.Transportation Research Record2270, 1 (Jan. 2012), 76–86. doi:10.3141/ 2270-10 Publisher: SAGE Publications Inc

    2012

  38. [45]

    Danny Munera, Johnny Aguirre, Natalia Gaviria Gomez, et al. 2021. IoT-based air quality monitoring systems for smart cities: A systematic mapping study. International Journal of Electrical and Computer Engineering11, 4 (2021), 3470

    2021

  39. [46]

    M. V. Narayana, Kranthi Kumar Rachvarapu, Devendra Jalihal, and Shiva Na- gendra S. M. 2023. Sens-BERT: Enabling Transferability and Re-calibration of Calibration Models for Low-cost Sensors under Reference Measurements Scarcity. doi:10.48550/arXiv.2309.13390 arXiv:2309.13390

    2023 doi arXiv

  40. [47]

    Griswold

    Nima Nikzad, Nakul Verma, Celal Ziftci, Elizabeth Bales, Nichole Quick, Piero Zappi, Kevin Patrick, Sanjoy Dasgupta, Ingolf Krueger, Tajana Šimunić Rosing, and William G. Griswold. 2012. CitiSense: improving geospatial environmental assessment of air quality using a wireless p...

    2012

  41. [48]

    Alam Noor, Kai Li, Eduardo Tovar, Pei Zhang, and Bo Wei. 2024. Fusion flow- enhanced graph pooling residual networks for unmanned aerial vehicles surveil- lance in day and night dual visions.Engineering Applications of Artificial Intelli- gence136 (2024), 108959

    2024

  42. [49]

    Yoshinari Shirai, Yasue Kishino, Futoshi Naya, and Yutaka Yanagisawa. 2016. Toward On-Demand Urban Air Quality Monitoring using Public Vehicles. InPro- ceedings of the 2nd International Workshop on Smart (SmartCities ’16). Association for Computing Machinery, New York, NY, USA...

    2016

  43. [50]

    Minxing Si, Ying Xiong, Shan Du, and Ke Du. 2020. Evaluation and calibration of a low-cost particle sensor in ambient conditions using machine-learning methods. 13, 4 (2020), 1693–1707. doi:10.5194/amt-13-1693-2020

    2020 doi

  44. [52]

    Dede Tarwidi, Sri Redjeki Pudjaprasetya, Didit Adytia, and Mochamad Apri

  45. [53]

    Wataru Tsujita, Akihito Yoshino, Hiroshi Ishida, and Toyosaka Moriizumi. 2005. Gas sensor network for air-pollution monitoring.Sensors and Actuators B: Chem- ical110, 2 (Oct. 2005), 304–311. doi:10.1016/j.snb.2005.02.008

    2005 doi

  46. [54]

    Emre Tunca, Eyüp Selim Köksal, and Sakine Çetin Taner. 2023. Calibrating UAV thermal sensors using machine learning methods for improved accuracy in agricultural applications. 133 (2023), 104804. doi:10.1016/j.infrared.2023.104804

    2023

  47. [55]

    Martine Van Poppel, Philipp Schneider, Jan Peters, Sinan Yatkin, Michel Ger- boles, Christina Matheeussen, Alena Bartonova, Silvije Davila, Marco Signorini, Matthias Vogt, Franck René Dauge, Jøran Solnes Skaar, and Rolf Haugen. 2023. SensEURCity: A multi-city air quality datas...

    2023 doi

  48. [56]

    Edwin Villanueva, Soledad Espezua, George Castelar, Kyara Diaz, and Erick Ingaroca. 2023. Smart Multi-Sensor Calibration of Low-Cost Particulate Matter Yin et al. Monitors. 23, 7 (2023), 3776. doi:10.3390/s23073776

    2023 doi

  49. [57]

    World Health Organization (WHO). n.d.. Air Pollution Data Portal. https: //www.who.int/data/gho/data/themes/air-pollution. Accessed: 2025-03-29

    2025

  50. [58]

    Zhengwei Wu, Xiaoxi Zhang, Susu Xu, Xinlei Chen, Pei Zhang, Hae Young Noh, and Carlee Joe-Wong. 2020. A generative simulation platform for multi-agent systems with incentives. InAdjunct Proceedings of the 2020 ACM International Joint Conference on Pervasive and Ubiquitous Comp...

    2020

  51. [59]

    Susu Xu, Xinlei Chen, Xidong Pi, Carlee Joe-Wong, Pei Zhang, and Hae Young Noh. 2019. Incentivizing vehicular crowdsensing system for large scale smart city applications. InSensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2019, Vol. 10970....

    2019 doi

  52. [60]

    Susu Xu, Xinlei Chen, Xidong Pi, Carlee Joe-Wong, Pei Zhang, and Hae Young Noh. 2019. Vehicle dispatching for sensing coverage optimization in mobile crowdsensing systems: poster abstract. InProceedings of the 18th International Conference on Information Processing in Sensor N...

    2019

  53. [61]

    Xiangxiang Xu, Xinlei Chen, Xinyu Liu, Hae Young Noh, Pei Zhang, and Lin Zhang. 2016. Gotcha II: Deployment of a Vehicle-based Environmental Sensing System: Poster Abstract. InProceedings of the 14th ACM Conference on Embedded Network Sensor Systems CD-ROM. ACM, Stanford CA US...

    2016

  54. [62]

    2025.Sur- vey of Computational and Sensing Methods for Pre-emergent Zoonotic Viral Pathogen Tracking

    Kevin Yin, Kevin Maffetone, Yahya Naveed, and Julia Gersey. 2025.Sur- vey of Computational and Sensing Methods for Pre-emergent Zoonotic Viral Pathogen Tracking. https://www.researchgate.net/publication/ 390027752_Survey_of_Computational_and_Sensing_Methods_for_Pre- emergent_Z...

    2025

  55. [2020]

    doi:10.1002/cpe.5179 _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/cpe.5179

    New technologies for smart cities – high-resolution air pollution maps based on intelligent sensors.Concurrency and Computation: Prac- tice and Experience32, 13 (2020), e5179. doi:10.1002/cpe.5179 _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/cpe.5179

    2020 doi

  56. [2022]

    InProceedings of the 23rd Annual International Workshop on Mobile Computing Systems and Applications

    A neural-based bandit approach to mobile crowdsourcing. InProceedings of the 23rd Annual International Workshop on Mobile Computing Systems and Applications. ACM, Tempe Arizona, 15–21. doi:10.1145/3508396.3512886

  57. [2023]

    10 (2023), 102119

    An optimized XGBoost-based machine learning method for predicting wave run-up on a sloping beach. 10 (2023), 102119. doi:10.1016/j.mex.2023.102119

    2023

  58. [2025]

    arXiv:2503.11727 [cs.CY] https://arxiv.org/abs/2503.11727

    Survey of City-Wide Homelessness Detection Through Environmental Sensing. arXiv:2503.11727 [cs.CY] https://arxiv.org/abs/2503.11727

    arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.