중소도시 미세먼지 농도 예측을 위한 머신러닝 기법 적용 가능성 평가 = Evaluation of Machine Learning Application on the Prediction of Particulate Matter Concentrations in Small/Medium-Sized City

국문 초록 (Abstract)

목적:본 연구는 중소도시의 대기환경 및 기상 데이터를 이용하여 미세먼지(PM10 및 PM2.5) 농도를 예측하는 머신러닝 모델의 성능을 평가하였다.
방법:Multiple Linear Regression (MLR), Decision Tree Regression (DTR), Random Forest (RF), Extreme Gradient Boosting (XGB), Light Gradient Boosting Machine (LGB)의 5개 머신러닝 모델을 이용하여 PM10과 PM2.5 농도를 예측하였다. 2017년부터 2022년까지 5개의 대기환경 데이터(NO2, SO2, CO, PM10, PM2.5)와 7개의 기상 데이터(기온, 습도, 증기압, 풍속, 강수량, 현지기압, 해면기압)를 3개의 대기질 측정소와 1개의 기상대에서 수집하였고, 이를 머신러닝 학습, 검증, 테스트 자료로 사용하였다. 머신러닝 예측 성능은 Root Mean Squared Error(RMSE), Mean Absolute Error(MAE), Coefficient of determination(R2)의 3가지 성능 평가 지표를 이용하여 정확도를 평가하였다.
결과 및 토의:미세먼지(PM10 및 PM2.5) 농도 예측에서 입력 변수에 예측하고자 하는 미세먼지 정보가 포함된 데이터셋을 학습한 모델이 그렇지 않은 모델보다 더 높은 예측 성능을 보였다. XGB 모델이 다른 머신러닝 모델들 보다 대부분의 경우에 가장 우수한 성능을 나타내었다. 그러나 미세먼지 농도를 예측하는 시간이 길어질수록 모든 머신러닝의 예측 정확도가 크게 감소하였다.
결론:본 연구에서 머신러닝 기법을 이용한 1시간 후의 미세먼지(PM10 및 PM2.5) 농도 예측은 충분히 적용 가능한 수준의 결과로 확인되었다. 그러나 장기간의 미세먼지를 예측하기 위해서는 딥러닝 모델을 적용하거나 미세먼지 농도에 영향을 미치는 추가적인 영향 인자 데이터를 확보하여 적용하는 것이 필요할 것이다.

다국어 초록 (Multilingual Abstract)

Objectives:In this study, Machine Learning (ML) algorithms were evaluated to predict the concentration of particulate matter (PM10 and PM2.5) using air quality and meteorological data in small/medium-sized city.
Methods:ML models, including Multiple Linear Regression (MLR), Decision Tree Regression (DTR), Random Forest (RF), Extreme Gradient Boosting (XGB), Light Gradient Boosting Machine (LGB), were used to predict PM10 and PM2.5 concentrations. Five air quality variables, including NO2, SO2, CO, PM10 and PM2.5, and seven meteorological variables, including temperature, humidity, vapor pressure, wind speed, precipitation, local atmospheric pressure, and sea-level atmosphere pressure, were collected from three air quality monitoring stations and one meteorological observatory from 2017 to 2022. A total of 52,583 sets of data were used for ML. The prediction accuracies of the applied ML models were evaluated using the Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Coefficient of determination (R2).
Results and Discussion:Higher ML performance was obtained when using the data including PM10 and PM2.5 compared to the data excluding these variables. Among five different ML models, the XGB model showed the highest accuracy in predicting PM10 and PM2.5 one hour in the future. However, poorer performance was obtained as the predicted period increased from one hour to 72 hours.
Conclusion:The application of ML algorithms for the short-term prediction of PM10 and PM2.5 was successful in this study. However, more input variables and deep learning algorithms are needed for long-term prediction of PM10 and PM2.5.

참고문헌 (Reference)

1 T. Chen, "XGBoost: a scalable tree boosting system, Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining" Association for Computing Machinery 785-794, 2016

2 A. Mukherjee, "World air particulate matter: sources, distribution and health effects" 15 (15): 283-309, 2017

3 "World Health Organization (WHO)"

4 F. Hosseinpour, "Using machine learning to improve the estimate of U.S. background ozone" 316 : 120145-, 2024

5 D. A. Wood, "Trend decomposition aids forecasts of air particulate matter(PM2. 5)assisted by machine and deep learning without recourse to exogenous data" 13 (13): 101352-, 2022

6 J. T. Pryor, "The physiological effects of air pollution: particulate matter, Physiology and Disease" 10 : 882569-, 2022

7 M. A. Cole, "The impact of the wuhan covid-19 lockdown on air pollution and health : a machine learning and augmented synthetic control approach" 76 (76): 553-580, 2020

8 D. Chicco, "The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation" 7 : e623-, 2021

9 Y. Xie, "Spatiotemporal variations of PM2.5 and PM10 concentrations between 31 chinese cities and their relationships with SO2, NO2, CO and O3" 20 : 141-149, 2015

10 Y. Zhan, "Spatiotemporal prediction of continuous daily PM2. 5 concentrations across china using a spatially explicit machine learning algorithm" 155 : 129-139, 2017

11 S. Araki, "Spatiotemporal land use random forest model for estimating metropolitan NO2 exposure in Japan" 634 : 1269-1277, 2018

12 B. Y. Kim, "Short-term prediction of particulate matter (PM10 and PM2.5) in seoul, South Korea using tree-based machine learning algorithms" 13 (13): 101547-, 2022

13 E. Kristiani, "Short-term prediction of PM2. 5 using LSTM deep learning methods" 14 (14): 2068-, 2022

14 T. Chai, "Root mean square error(RMSE)or mean absolute error(MAE)?-Arguments against avoiding RMSE in the literature" 7 (7): 1247-1250, 2014

15 L. Li, "Recent advances in artificial intelligence and machine learning for nonlinear relationship analysis and process control in drinking water treatment : a review" 405 : 126673-, 2021

16 L. Breiman, "Random forests" 45 (45): 5-32, 2001

17 박정은 ; 조재훈 ; 김용태, "Prediction of PM2. 5 dual-stage attention mechanism using surrounding area data" 32 (32): 416-423, 2022

18 A. Barthwal, "Prediction and analysis of particulate matter(PM2. 5 and PM10)concentrations using machine learning techniques" 14 (14): 1323-1338, 2023

19 C. Brokamp, "Predicting daily urban fine particulate matter concentrations using a random forest model" 52 (52): 4173-4179, 2018

20 Z. Gao, "Predicting PM2. 5 levels and exceedance days using machine learning methods" 323 : 120396-, 2024

21 S. U. Kim, "Physical and chemical mechanisms of the daily-to-seasonal variation of PM10 in Korea" 712 : 136429-, 2020

22 M. Rahimzad, "Performance comparison of an LSTM-based deep learning model versus conventional machine learning algorithms for streamflow forecasting" 35 (35): 4167-4187, 2021

23 J. Lee, "Particulate matter exposure and neurodegenerative diseases: a comprehensive update on toxicity and mechanisms" 115565-, 2023

24 R. B. Hamanaka, "Particulate matter air pollution:effects on the cardiovascular system" 9 : 680-, 2018

25 W. Zhang, "Modeling, optimization and understanding of adsorption process for pollutant removal via machine learning: recent progress and future perspectives" 311 : 137044-, 2023

26 D. Xiao, "Machine learning-based rapid response tools for regional air pollution modelling" 199 : 463-473, 2019

27 Y. Chi, "Machine learning-based estimation of ground-level NO2concentrations over China" 807 : 150721-, 2022

28 C. Janiesch, "Machine learning and deep learning" 31 (31): 685-695, 2021

29 X. Su, "Linear regression" 4 (4): 275-294, 2012

30 G. Ke, "Lightgbm: a highly efficient gradient boosting decision tree" 30 : 2017

31 최종규 ; 최인순 ; 조광근 ; 이승호, "Harmfulness of particulate matter in disease progression" 30 (30): 191-201, 2020

32 P. Sadorsky, "Forecasting solar stock prices using tree-based machine learning classification : how important are silver prices?" 61 : 101705-, 2022

33 J. Du, "Forecasting ground-level ozone concentration levels using machine learning, Resources" 184 : 106380-, 2022

34 J. Zhao, "Forecasting fine particulate matter concentrations by in-depth learning model according to random forest and bilateral long-and short-term memory neural networks" 14 (14): 9430-, 2022

35 L. Mampitiya, "Forecasting PM10 levels in sri lanka : a comparative analysis of machine learning models PM10" 13 : 100395-, 2024

36 J. S. Pérez-Carrasquilla, "Forecasting 24h averaged PM2.5concentration in the Aburrá Valley using tree-based machine learning models, global forecasts, and satellite information" 9 (9): 121-135, 2023

37 Y. Kang, "Estimation of surface-level NO2 and O3 concentrations using TROPOMI data and machine learning over east asia" 288 : 117711-, 2021

38 Y. Yang, "Estimation of PM2. 5 concentration across china based on multi-source remote sensing data and machine learning methods" 16 (16): 467-, 2024

39 X. Hu, "Estimating PM2. 5 concentrations in the conterminous united states using the random forest approach" 51 (51): 6936-6944, 2017

40 C. C. Wei, "Establishing a real-time prediction system for fine particulate matter concentration using machine-learning models" 14 (14): 1817-, 2023

41 S. X. Lv, "Effective machine learning model combination based on selective ensemble strategy for time series forecasting" 612 : 994-1023, 2022

42 B. Zhang, "Deep learning for air pollutant concentration prediction : a review" 290 : 119347-, 2022

43 Y. Ren, "Deep learning coupled model based on TCN-LSTM for particulate matter concentration prediction" 14 (14): 101703-, 2023

44 S. Buschjäger, "Decision tree and random forest implementations for fast filtering of sensor data" 65 (65): 209-222, 2018

45 X. Liu, "Data-driven machine learning in environmental pollution : gains and problems" 56 (56): 2124-2133, 2022

46 D. M. Zhang, "Data-and experience-driven neural networks for long-term settlement prediction of tunnel" 147 : 105669-, 2024

47 A. D. Bucchianico, "Coefficient of determination (R2)"

48 B. Czernecki, "Assessment of machine learning algorithms in short-term forecasting of PM10and PM2. 5 concentrations in selected polish agglomerations" 21 (21): 200586-, 2021

49 M. Sharma, "Assessment of fine particulate matter for port city of eastern peninsular india using gradient boosting machine learning model" 13 (13): 743-, 2022

50 Z. Peng, "Application of machine learning in atmospheric pollution research : a state-of-art review" 910 : 168588-, 2024

51 L. Lv, "Application of machine learning algorithms to improve numerical simulation prediction of PM2. 5and chemical components" 12 (12): 101211-, 2021

52 National Institute of Environmental Research(NIER), "Annual Report of Air Quality in Korea" 1-403, 2023

53 최우철 ; 정규수, "Analysis of the factors affecting fine dust concentration before and after COVID-19" 21 (21): 395-402, 2021

54 A. Bekkar, "Air-pollution prediction in smart city, deep learning approach" 8 (8): 161-, 2021

55 B. Panneerselvam, "A novel approach for the prediction and analysis of daily concentrations of particulate matter using machine learning" 897 : 166178-, 2023

56 Z. Zhang, "A hybrid deep learning technology for PM2. 5 air quality forecasting" 28 (28): 39409-39422, 2021

57 A. M. Ahmed, "A decision tree algorithm combined with linear regression for data classification" 12-14, 2018

58 A. Pandya, "A comparative and systematic study of machine learning(ML)approaches for particulate matter(PM)prediction" 31 (31): 595-614, 2024