Calibration of Low-Cost LoRaWAN-Based IoT Air Quality Monitors Using the Super Learner Ensemble: A Case Study for Accurate Particulate Matter Measurement
Abstract
:1. Introduction
2. Materials and Methods
2.1. Air Quality Monitors Utilized in the Calibration Study
Research-Grade Air Quality Monitor: Palas Fidas Frog
- PM1.0: Particulate matter with an aerodynamic diameter ≤ 1.0 m (unit is g/m3).
- PM2.5: Particulate matter with an aerodynamic diameter ≤ 2.5 m (unit is g/m3).
- PM4.0: Particulate matter with an aerodynamic diameter ≤ 4.0 m (unit is g/m3).
- PM10.0: Particulate matter with an aerodynamic diameter ≤ 10.0 m (unit is g/m3).
- Total PM Concentration: The overall concentration of particulate matter across different size fractions (unit is g/m3).
- Particle Count Density: The number of particles per unit volume of air (unit is number of particles/cm3 or #/cm3).
2.2. Low-Cost Air Quality Monitor: LoRaWAN Prototype
- PPD42NS: Particle CounterThe PPD42NS (Figure 2a) [23] is the primary PM sensor in the LoRa node. It is an affordable optical sensor designed to measure parameters associated with particulate matter (PM) concentrations for two size ranges, detected via two separate channels. Channel 1 measures parameters associated with particulates larger than 1 m in diameter, while Channel 2 measures parameters associated with particulates larger than 2.5 m in diameter. The measurement range for PM concentration is approximately from 0 to 28,000 g/m3 for both >1 m (P1) and >2.5 m (P2) particle sizes.The sensor operates based on the Low Pulse Occupancy (LPO) principle. When particles pass through the sensor’s optical chamber, they scatter the light emitted by an LED. This scattered light is detected by a photodiode, which outputs a pulse width-modulated (PWM) signal. The duration of the PWM signal is proportional to the particle count and size. LPO is defined as the amount of time during which this PWM signal is low during a fixed sampling interval (15 s in this study). The standard error for LPO is approximately 0.02 units for both channels under low-concentration conditions (<50 g/m3). Under high-concentration conditions (≥50 g/m3), the error varies nonlinearly [24,25].The following parameters are measured by the PPD42NS:
- -
- P1 LPO: Represents the total time for Channel 1 (indicating the presence of particles larger than 1 m) during which the sensor signal is low in a 15 s sampling period. It is also referred to as the >1 μm LPO and is measured in milliseconds.
- -
- P1 ratio: Represents the proportion of time the sensor signal is low for Channel 1 (indicating the presence of particles larger than 1 m) during the sampling period. It is also referred to as the >1 μm ratio.
- -
- P1 concentration: Measures the PM concentration of particles larger than 1 m in diameter. It is also referred to as the >1 μm concentration and is measured in g/m3.
- -
- P2 LPO: Represents the total time for Channel 2 (indicating the presence of particles larger than 2.5 m) during which the sensor signal is low in a 15 s sampling period. It is also referred to as the >2.5 μm LPO and is measured in milliseconds.
- -
- P2 ratio: Represents the proportion of time the sensor signal is low for Channel 2 (indicating the presence of particles larger than 2.5 m) during the sampling period. It is also referred to as the >2.5 μm ratio.
- -
- P2 concentration: Measures the PM concentration of particles larger than 2.5 m in diameter. It is also referred to as the >2.5 μm concentration and is measured in g/m3.
- BME280: Climate SensorThe BME280 (Figure 2b) is the climate sensor used in the LoRa node. It measures air temperature (referred to as temperature), atmospheric pressure (referred to as pressure), and relative humidity (referred to as humidity), which are critical for understanding environmental conditions and calibrating the PM sensor. The BME280 sensor can measure temperatures ranging from −40 °C to 85 °C with an accuracy of ±0.5 °C, pressure from 300 hPa to 1100 hPa with an accuracy of ±1.0 hPa, and humidity from 0% to 100% with an accuracy of ±3% [26].
2.3. Air Quality Monitoring Data Collection
2.4. Supervised ML Approach for Calibration
Super Learner
2.5. Metric for ML Calibration
- is the actual target value for the i-th data point.
- is the predicted target value for the i-th data point.
- is the mean of all target values.
- N is the total number of data points.
2.6. Hyperparameter Tuning for ML Models
2.7. Feature Ranking for ML Calibration
2.8. Basic ML Calibration Work Flow
2.9. Preprocessing of LoRa and Palas Sensor Data
3. Results
4. Challenges and Future Work
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Abbreviations
General Technical Terms | |
CSV | Comma-Separated Values |
FEM | Federal Equivalent Method |
IoT | Internet of Things |
LoRaWAN | Long Range Wide Area Network |
LPO | Low Pulse Occupancy |
Probability Density Function | |
PM | Particulate Matter |
PM1.0 | Particulate Matter with an aerodynamic diameter ≤ 1.0 m |
PM2.5 | Particulate Matter with an aerodynamic diameter ≤ 2.5 m |
PM4.0 | Particulate Matter with an aerodynamic diameter ≤ 4.0 m |
PM10.0 | Particulate Matter with an aerodynamic diameter ≤ 10.0 m |
PWM | Pulse Width Modulation |
Quantile—Quantile | |
R2 | R-squared or Coefficient of Determination |
UTD | University of Texas at Dallas |
WLAN | Wireless Local Area Network |
WSTC | Waterview Science and Technology Center |
Model and Method-Related Terms | |
DT | Decision Tree |
EB | Ensemble Bagging |
KNN | K-Nearest Neighbors |
LightGBM | Light Gradient Boosting Machine |
LR | Linear Regression |
NN | Neural Networks |
RF | Random Forest |
RR | Ridge Regression |
SL | Super Learner |
XGBoost | Extreme Gradient Boosting |
References
- Zimmerman, N.; Presto, A.A.; Kumar, S.P.; Gu, J.; Hauryliuk, A.; Robinson, E.S.; Robinson, A.L.; Subramanian, R. A machine learning calibration model using random forests to improve sensor performance for lower-cost air quality monitoring. Atmos. Meas. Tech. 2018, 11, 291–313. [Google Scholar] [CrossRef]
- Alfano, B.; Barretta, L.; Del Giudice, A.; De Vito, S.; Di Francia, G.; Esposito, E.; Formisano, F.; Massera, E.; Miglietta, M.L.; Polichetti, T. A review of low-cost particulate matter sensors from the developers’ perspectives. Sensors 2020, 20, 6819. [Google Scholar] [CrossRef] [PubMed]
- Concas, F.; Mineraud, J.; Lagerspetz, E.; Varjonen, S.; Liu, X.; Puolamäki, K.; Nurmi, P.; Tarkoma, S. Low-cost outdoor air quality monitoring and sensor calibration: A survey and critical analysis. ACM Trans. Sens. Netw. (TOSN) 2021, 17, 1–44. [Google Scholar] [CrossRef]
- Clarity. Low-Cost Sensors for Comprehensive Air Quality Management. 2025. Available online: https://www.clarity.io/blog/low-cost-sensors-for-comprehensive-air-quality-management (accessed on 31 January 2025).
- Caldas, F.; Soares, C. Machine learning in orbit estimation: A survey. Acta Astronaut. 2024, 220, 97–107. [Google Scholar] [CrossRef]
- Styp-Rekowski, K.; Michaelis, I.; Stolle, C.; Baerenzung, J.; Korte, M.; Kao, O. Machine learning-based calibration of the GOCE satellite platform magnetometers. Earth Planets Space 2022, 74, 138. [Google Scholar] [CrossRef]
- Wijeratne, L.O.H. Coupling Physical Measurement with Machine Learning for Holistic Environmental Sensing. Ph.D. Thesis, University of Texas at Dallas, Dallas, TX, USA, 2021. Available online: https://utd-ir.tdl.org/items/5550059c-4af4-4da8-97a3-c5050baec902 (accessed on 31 January 2025).
- Wijeratne, L.O.; Kiv, D.R.; Aker, A.R.; Talebi, S.; Lary, D.J. Using machine learning for the calibration of airborne particulate sensors. Sensors 2019, 20, 99. [Google Scholar] [CrossRef]
- Zhang, Y.; Wijeratne, L.O.; Talebi, S.; Lary, D.J. Machine learning for light sensor calibration. Sensors 2021, 21, 6259. [Google Scholar] [CrossRef]
- Park, D.; Yoo, G.W.; Park, S.H.; Lee, J.H. Assessment and calibration of a low-cost PM2.5 sensor using machine learning (HybridLSTM neural network): Feasibility study to build an air quality monitoring system. Atmosphere 2021, 12, 1306. [Google Scholar] [CrossRef]
- Naimi, A.I.; Balzer, L.B. Stacked generalization: An introduction to super learning. Eur. J. Epidemiol. 2018, 33, 459–464. [Google Scholar] [CrossRef]
- de Carvalho Silva, J.; Rodrigues, J.J.; Alberti, A.M.; Solic, P.; Aquino, A.L. LoRaWAN—A low power WAN protocol for Internet of Things: A review and opportunities. In Proceedings of the 2017 2nd International Multidisciplinary Conference on Computer and Energy Science (SpliTech), Split, Croatia, 12–14 July 2017; IEEE: Piscataway, NJ, USA, 2017; pp. 1–6. [Google Scholar]
- Haxhibeqiri, J.; De Poorter, E.; Moerman, I.; Hoebeke, J. A survey of LoRaWAN for IoT: From technology to application. Sensors 2018, 18, 3995. [Google Scholar] [CrossRef]
- Adelantado, F.; Vilajosana, X.; Tuset-Peiro, P.; Martinez, B.; Melia-Segui, J.; Watteyne, T. Understanding the Limits of LoRaWAN. IEEE Commun. Mag. 2017, 55, 34–40. [Google Scholar] [CrossRef]
- Thu, M.Y.; Htun, W.; Aung, Y.L.; Shwe, P.E.E.; Tun, N.M. Smart air quality monitoring system with LoRaWAN. In Proceedings of the 2018 IEEE International Conference on Internet of Things and Intelligence System (IOTAIS), Bali, Indonesia, 1–3 November 2018; IEEE: Piscataway, NJ, USA, 2018; pp. 10–15. [Google Scholar]
- Bonilla, V.; Campoverde, B.; Yoo, S.G. A Systematic Literature Review of LoRaWAN: Sensors and Applications. Sensors 2023, 23, 8440. [Google Scholar] [CrossRef] [PubMed]
- Palas GmbH. Fidas Frog Product Datasheet. 2025. Available online: https://www.palas.de/en/product/download/fidasfrog/datasheet/pdf (accessed on 2 January 2025).
- Palas GmbH. Palas Technical Feature. 2025. Available online: https://www.palas.de/en/product/fidas200 (accessed on 11 January 2015).
- Lock, J.A.; Gouesbet, G. Generalized Lorenz—Mie theory and applications. J. Quant. Spectrosc. Radiat. Transf. 2009, 110, 800–807. [Google Scholar] [CrossRef]
- Seeed Studio. Grove—Dust Sensor (PPD42NS). 2025. Available online: https://www.seeedstudio.com/Grove-Dust-Sensor-PPD42NS.html (accessed on 23 February 2025).
- Arduino Store. Grove—Temp&Humi&Barometer Sensor (BME280). 2025. Available online: https://store-usa.arduino.cc/products/grove-temp-humi-barometer-sensor-bme280 (accessed on 23 February 2025).
- Seeed Studio Wiki. Grove—Barometer Sensor (BME280). 2025. Available online: https://wiki.seeedstudio.com/Grove-Barometer_Sensor-BME280/ (accessed on 23 February 2025).
- Seeed Studio. Grove Dust Sensor Product Datasheet. 2025. Available online: https://www.mouser.com/datasheet/2/744/Seeed_101020012-1217636.pdf (accessed on 11 January 2025).
- Canu, M.; Galvis, B.; Morales, R.; Ramírez, O.; Madelin, M. Understanding the Shinyei PPD24NS Low-Cost Dust Sensor. In Proceedings of the 2018 IEEE International Conference on Environmental Engineering, Milan, Italy, 12–14 March 2018. [Google Scholar]
- Austin, E.; Novosselov, I.; Seto, E.; Yost, M.G. Laboratory Evaluation of the Shinyei PPD42NS Low-Cost Particulate Matter Sensor. PLoS ONE 2015, 10, e0137789. [Google Scholar] [CrossRef]
- Bosch Sensortec. BME280: Combined Humidity and Pressure Sensor Datasheet. 2025. Available online: https://www.bosch-sensortec.com/media/boschsensortec/downloads/datasheets/bst-bme280-ds002.pdf (accessed on 11 January 2025).
- El Naqa, I.; Murphy, M.J. What is Machine Learning; Springer: Berlin/Heidelberg, Germany, 2015. [Google Scholar] [CrossRef]
- Sykes, A.O. An Introduction to Regression Analysis. Coase-Sandor Inst. Law Econ. Work. Pap. 1993, 20, 1–39. Available online: https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?article=1050&context=law_and_economics (accessed on 23 February 2025).
- Ruwali, S.; Fernando, B.; Talebi, S.; Wijeratne, L.; Waczak, J.; Madusanka, P.M.; Lary, D.J.; Sadler, J.; Lary, T.; Lary, M.; et al. Estimating Inhaled Nitrogen Dioxide from the Human Biometric Response. Adv. Environ. Eng. Res. 2024, 5, 1–12. [Google Scholar] [CrossRef]
- Ruwali, S.; Fernando, B.A.; Talebi, S.; Wijeratne, L.; Waczak, J.; Sooriyaarachchi, V.; Hathurusinghe, P.; Lary, D.J.; Sadler, J.; Lary, T.; et al. Gauging ambient environmental carbon dioxide concentration solely using biometric observations: A machine learning approach. Med. Res. Arch. 2024, 12. [Google Scholar] [CrossRef]
- Ruwali, S.; Talebi, S.; Fernando, A.; Wijeratne, L.O.; Waczak, J.; Dewage, P.M.; Lary, D.J.; Sadler, J.; Lary, T.; Lary, M.; et al. Quantifying Inhaled Concentrations of Particulate Matter, Carbon Dioxide, Nitrogen Dioxide, and Nitric Oxide Using Observed Biometric Responses with Machine Learning. BioMedInformatics 2024, 4, 1019–1046. [Google Scholar] [CrossRef]
- Adong, P.; Bainomugisha, E.; Okure, D.; Sserunjogi, R. Applying machine learning for large scale field calibration of low-cost PM2.5 and PM10 air pollution sensors. Appl. AI Lett. 2022, 3, e76. [Google Scholar] [CrossRef]
- Si, M.; Xiong, Y.; Du, S.; Du, K. Evaluation and calibration of a low-cost particle sensor in ambient conditions using machine-learning methods. Atmos. Meas. Tech. 2020, 13, 1693–1707. [Google Scholar] [CrossRef]
- Kiss, L.M.; Pintér, J.M.; Kósa, B.; Markó, B.; Veres, L. Calibration of particle sensor with neural network algorithm. Pollack Period. 2024, 19. [Google Scholar] [CrossRef]
- Liu, B.; Tan, X.; Jin, Y.; Yu, W.; Li, C. Application of RR-XGBoost combined model in data calibration of micro air quality detector. Sci. Rep. 2021, 11, 15662. [Google Scholar] [CrossRef]
- Phillips, R.V.; Van Der Laan, M.J.; Lee, H.; Gruber, S. Practical considerations for specifying a super learner. Int. J. Epidemiol. 2023, 52, 1276–1285. [Google Scholar] [CrossRef] [PubMed]
- Dey, R.; Mathur, R. Ensemble learning method using stacking with base learner, a comparison. In Proceedings of the International Conference on Data Analytics and Insights, Kolkata, WIndia, 11–13 May 2023; Springer: Berlin/Heidelberg, Germany, 2023; pp. 159–169. [Google Scholar] [CrossRef]
- Polley, E.C.; Van der Laan, M.J. Super Learner in Prediction; University of California, Berkeley: Berkeley, CA, USA, 2010; Available online: https://biostats.bepress.com/ucbbiostat/paper266/?TB_iframe=true&width=370.8&height=658 (accessed on 31 January 2025).
- Wong, J.; Manderson, T.; Abrahamowicz, M.; Buckeridge, D.L.; Tamblyn, R. Can hyperparameter tuning improve the performance of a super learner?: A case study. Epidemiology 2019, 30, 521–531. [Google Scholar] [CrossRef] [PubMed]
- Ladds, M.A.; Thompson, A.P.; Kadar, J.P.; Slip, D.J.; Hocking, D.P.; Harcourt, R.G. Super Machine Learning: Improving Accuracy and Reducing Variance of Behaviour Classification from Accelerometry. Anim. Biotelemetry 2017, 5, 8. [Google Scholar] [CrossRef]
- Ozer, D.J. Correlation and the Coefficient of Determination. Psychol. Bull. 1985, 97, 307. Available online: https://psycnet.apa.org/fulltext/1985-19147-001.html (accessed on 31 January 2025). [CrossRef]
- Bergstra, J.; Bengio, Y. Random Search for Hyper-Parameter Optimization. J. Mach. Learn. Res. 2012, 13, 281–305. Available online: http://jmlr.org/papers/v13/bergstra12a.html (accessed on 31 January 2025).
- Altmann, A.; Toloşi, L.; Sander, O.; Lengauer, T. Permutation Importance: A Corrected Feature Importance Measure. Bioinformatics 2010, 26, 1340–1347. [Google Scholar] [CrossRef]
- Zender-Świercz, E.; Galiszewska, B.; Telejko, M.; Starzomska, M. The effect of temperature and humidity of air on the concentration of particulate matter-PM2.5 and PM10. Atmos. Res. 2024, 312, 107733. [Google Scholar] [CrossRef]
- Kim, M.; Jeong, S.G.; Park, J.; Kim, S.; Lee, J.H. Investigating the impact of relative humidity and air tightness on PM sedimentation and concentration reduction. Build. Environ. 2023, 241, 110270. [Google Scholar] [CrossRef]
- Lou, C.; Liu, H.; Li, Y.; Peng, Y.; Wang, J.; Dai, L. Relationships of relative humidity with PM 2.5 and PM 10 in the Yangtze River Delta, China. Environ. Monit. Assess. 2017, 189, 582. [Google Scholar] [CrossRef]
- Zhang, L.; Cheng, Y.; Zhang, Y.; He, Y.; Gu, Z.; Yu, C. Impact of air humidity fluctuation on the rise of PM mass concentration based on the high-resolution monitoring data. Aerosol Air Qual. Res. 2017, 17, 543–552. [Google Scholar] [CrossRef]
- Haikerwal, A.; Akram, M.; Del Monaco, A.; Smith, K.; Sim, M.R.; Meyer, M.; Tonkin, A.M.; Abramson, M.J.; Dennekamp, M. Impact of fine particulate matter (PM 2.5) exposure during wildfires on cardiovascular health outcomes. J. Am. Heart Assoc. 2015, 4, e001653. [Google Scholar] [CrossRef]
- Péré, J.; Bessagnet, B.; Mallet, M.; Waquet, F.; Chiapello, I.; Minvielle, F.; Pont, V.; Menut, L. Direct radiative effect of the Russian wildfires and its impact on air temperature and atmospheric dynamics during August 2010. Atmos. Chem. Phys. 2014, 14, 1999–2013. [Google Scholar] [CrossRef]
- Shaposhnikov, D.; Revich, B.; Bellander, T.; Bedada, G.B.; Bottai, M.; Kharkova, T.; Kvasha, E.; Lezina, E.; Lind, T.; Semutnikova, E.; et al. Mortality related to air pollution with the Moscow heat wave and wildfire of 2010. Epidemiology 2014, 25, 359–364. [Google Scholar] [CrossRef]
- Alvarez, H.B.; Echeverria, R.S.; Alvarez, P.S.; Krupa, S. Air quality standards for particulate matter (PM) at high altitude cities. Environ. Pollut. 2013, 173, 255–256. [Google Scholar] [CrossRef]
- Chen, Z.; Cheng, S.; Li, J.; Guo, X.; Wang, W.; Chen, D. Relationship between atmospheric pollution processes and synoptic pressure patterns in northern China. Atmos. Environ. 2008, 42, 6078–6087. [Google Scholar] [CrossRef]
- Li, H.; Guo, B.; Han, M.; Tian, M.; Zhang, J. Particulate matters pollution characteristic and the correlation between PM (PM2.5, PM10) and meteorological factors during the summer in Shijiazhuang. J. Environ. Prot. 2015, 6, 457–463. [Google Scholar] [CrossRef]
- Czernecki, B.; Półrolniczak, M.; Kolendowicz, L.; Marosz, M.; Kendzierski, S.; Pilguj, N. Influence of the atmospheric conditions on PM10 concentrations in Poznań, Poland. J. Atmos. Chem. 2017, 74, 115–139. [Google Scholar] [CrossRef]
- Kaiser Permanente. Air Quality Program Expands with Kaiser Permanente Grant. 2025. Available online: https://about.kaiserpermanente.org/news/press-release-archive/air-quality-program-expands-with-kaiser-permanente-grant (accessed on 31 January 2025).
- New York City Department of Health and Mental Hygiene. Air Quality: NYC Community Air Survey (NYCCAS). 2025. Available online: https://www.nyc.gov/site/doh/data/data-sets/air-quality-nyc-community-air-survey.page (accessed on 31 January 2025).
- Breathe London. Breathe London: Mapping Air Pollution in Real-Time. 2025. Available online: https://www.breathelondon.org/ (accessed on 31 January 2025).
- Chu, J. A New Approach Could Change How We Track Extreme Air Pollution Events. 2021. Available online: https://news.mit.edu/2021/new-approach-could-change-how-we-track-extreme-air-pollution-events-0630 (accessed on 31 January 2025).
- Rika Sensors. The Importance of Air Quality Sensors. 2025. Available online: https://www.rikasensor.com/blog-the-importance-of-air-quality-sensors.html (accessed on 31 January 2025).
- Jayaratne, R.; Liu, X.; Thai, P.; Dunbabin, M.; Morawska, L. The influence of humidity on the performance of a low-cost air particle mass sensor and the effect of atmospheric fog. Atmos. Meas. Tech. 2018, 11, 4883–4890. [Google Scholar] [CrossRef]
- Papadimitriou, C.H. Computational complexity. In Encyclopedia of Computer Science; John Wiley and Sons Ltd.: Chichester, UK, 2003; pp. 260–265. Available online: https://dl.acm.org/doi/abs/10.5555/1074100.1074233 (accessed on 31 January 2025).
- Chou, Y. R Parallel Processing for Developing Ensemble Learning with SuperLearner. 2019. Available online: https://yungchou.wordpress.com/2019/01/16/r-parallel-processing-for-developing-ensemble-learning-with-superlearner/ (accessed on 31 January 2025).
- UNICEF. Sustainable Development Goals. 2025. Available online: https://www.unicef.org/sustainable-development-goals (accessed on 30 January 2025).
- United Nations. Sustainable Development Goal 3: Ensure Healthy Lives and Promote Well-Being for All at All Ages. 2025. Available online: https://sdgs.un.org/goals/goal3 (accessed on 31 January 2025).
- Clarity Movement Co. The Myriad Use Cases for Low-Cost Air Quality Sensors. 2025. Available online: https://www.clarity.io/blog/the-myriad-use-cases-for-low-cost-air-quality-sensors (accessed on 30 January 2025).
- United Nations. Sustainable Development Goal 7: Ensure Access to Affordable, Reliable, Sustainable and Modern Energy for All. 2025. Available online: https://sdgs.un.org/goals/goal7 (accessed on 31 January 2025).
- United Nations. Sustainable Development Goal 11: Make Cities and Human Settlements Inclusive, Safe, Resilient, and Sustainable. 2025. Available online: https://sdgs.un.org/goals/goal11 (accessed on 31 January 2025).
- United Nations. Sustainable Development Goal 9: Build Resilient Infrastructure, Promote Inclusive and Sustainable Industrialization and Foster Innovation. 2025. Available online: https://sdgs.un.org/goals/goal9 (accessed on 31 January 2025).
Specification | Palas Fidas Frog | LoRaWAN Prototype |
---|---|---|
Measurement Range (PM) | 0–100 mg/m3 | 0–28,000 g/m3 |
Particle Size Range | 0.18–93 m | >1 m, >2.5 m |
Measurement Uncertainty | 9.7% (PM2.5), 7.5% (PM10) | 2% |
Power Source | Battery-operated | Solar-powered |
Cost | Approximately $20,000 USD | $100–$200 USD |
Model | PM1.0 | PM2.5 | PM4.0 | PM10.0 | Total PM Conc. | Particle Count Density | Avg. R2 | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Train | Test | Train | Test | Train | Test | Train | Test | Train | Test | Train | Test | Train | Test | |
Random Forest | 1.00 | 0.97 | 1.00 | 0.98 | 1.00 | 0.97 | 0.99 | 0.89 | 0.97 | 0.80 | 1.00 | 0.99 | 0.99 | 0.93 |
Ensemble Bagging | 1.00 | 0.96 | 1.00 | 0.97 | 1.00 | 0.97 | 0.98 | 0.90 | 0.96 | 0.80 | 1.00 | 0.99 | 0.99 | 0.93 |
Light Gradient Boosting Machine | 1.00 | 0.95 | 1.00 | 0.95 | 0.99 | 0.96 | 0.96 | 0.90 | 0.92 | 0.80 | 1.00 | 0.98 | 0.98 | 0.92 |
Extreme Gradient Boosting | 1.00 | 0.97 | 1.00 | 0.98 | 1.00 | 0.96 | 0.99 | 0.87 | 0.99 | 0.71 | 1.00 | 0.99 | 1.00 | 0.91 |
Decision Tree | 1.00 | 0.97 | 1.00 | 0.98 | 1.00 | 0.96 | 1.00 | 0.81 | 1.00 | 0.67 | 1.00 | 0.99 | 1.00 | 0.90 |
Neural Network | 0.95 | 0.88 | 0.94 | 0.88 | 0.93 | 0.88 | 0.63 | 0.59 | 0.45 | 0.40 | 0.95 | 0.92 | 0.81 | 0.76 |
K-Nearest Neighbors | 0.97 | 0.86 | 0.96 | 0.86 | 0.96 | 0.87 | 0.82 | 0.59 | 0.73 | 0.46 | 0.96 | 0.90 | 0.90 | 0.76 |
Linear Regression | 0.48 | 0.46 | 0.51 | 0.50 | 0.56 | 0.55 | 0.16 | 0.17 | 0.21 | 0.21 | 0.40 | 0.38 | 0.39 | 0.38 |
Ridge Regression | 0.48 | 0.46 | 0.51 | 0.50 | 0.56 | 0.55 | 0.15 | 0.17 | 0.21 | 0.21 | 0.40 | 0.38 | 0.39 | 0.38 |
Target Variable | Base Learners | Meta Learner | R2 Train | R2 Test |
---|---|---|---|---|
PM1.0 | Linear Regression, K-Nearest Neighbors, Extreme Gradient Boosting, Neural Network | Random Forest | 1.00 | 0.99 |
PM2.5 | K-Nearest Neighbors, Decision Tree, Bagging Regressor, Neural Network | Random Forest | 1.00 | 0.99 |
PM4.0 | Linear Regression, K-Nearest Neighbors, Decision Tree, Extreme Gradient Boosting | Random Forest | 1.00 | 0.99 |
PM10.0 | K-Nearest Neighbors, Random Forest, Extreme Gradient Boosting, Light Gradient Boosting Machine | Neural Network | 0.98 | 0.91 |
Total PM Concentration | K-Nearest Neighbors, Random Forest, Extreme Gradient Boosting, Light Gradient Boosting Machine | Neural Network | 0.97 | 0.86 |
Particle Count Density | Linear Regression, Ridge Regression, Decision Tree, Light Gradient Boosting Machine | Random Forest | 1.00 | 0.99 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Balagopal, G.; Wijeratne, L.; Waczak, J.; Hathurusinghe, P.; Iqbal, M.; Kiv, D.; Aker, A.; Lee, S.; Agnihotri, V.; Simmons, C.; et al. Calibration of Low-Cost LoRaWAN-Based IoT Air Quality Monitors Using the Super Learner Ensemble: A Case Study for Accurate Particulate Matter Measurement. Sensors 2025, 25, 1614. https://doi.org/10.3390/s25051614
Balagopal G, Wijeratne L, Waczak J, Hathurusinghe P, Iqbal M, Kiv D, Aker A, Lee S, Agnihotri V, Simmons C, et al. Calibration of Low-Cost LoRaWAN-Based IoT Air Quality Monitors Using the Super Learner Ensemble: A Case Study for Accurate Particulate Matter Measurement. Sensors. 2025; 25(5):1614. https://doi.org/10.3390/s25051614
Chicago/Turabian StyleBalagopal, Gokul, Lakitha Wijeratne, John Waczak, Prabuddha Hathurusinghe, Mazhar Iqbal, Daniel Kiv, Adam Aker, Seth Lee, Vardhan Agnihotri, Christopher Simmons, and et al. 2025. "Calibration of Low-Cost LoRaWAN-Based IoT Air Quality Monitors Using the Super Learner Ensemble: A Case Study for Accurate Particulate Matter Measurement" Sensors 25, no. 5: 1614. https://doi.org/10.3390/s25051614
APA StyleBalagopal, G., Wijeratne, L., Waczak, J., Hathurusinghe, P., Iqbal, M., Kiv, D., Aker, A., Lee, S., Agnihotri, V., Simmons, C., & Lary, D. J. (2025). Calibration of Low-Cost LoRaWAN-Based IoT Air Quality Monitors Using the Super Learner Ensemble: A Case Study for Accurate Particulate Matter Measurement. Sensors, 25(5), 1614. https://doi.org/10.3390/s25051614