Browse > Article
http://dx.doi.org/10.9717/kmms.2021.24.10.1414

Analysis of Input Factors and Performance Improvement of DNN PM2.5 Forecasting Model Using Layer-wise Relevance Propagation  

Yu, SukHyun (Dept. of Information, Electrical & Electronic Eng., Anyang University)
Publication Information
Journal of Korea Multimedia Society / v.24, no.10, 2021 , pp. 1414-1424 More about this Journal
Abstract
In this paper, the importance of input factors of a DNN (Deep Neural Network) PM2.5 forecasting model using LRP(Layer-wise Relevance Propagation) is analyzed, and forecasting performance is improved. Input factor importance analysis is performed by dividing the learning data into time and PM2.5 concentration. As a result, in the low concentration patterns, the importance of weather factors such as temperature, atmospheric pressure, and solar radiation is high, and in the high concentration patterns, the importance of air quality factors such as PM2.5, CO, and NO2 is high. As a result of analysis by time, the importance of the measurement factors is high in the case of the forecast for the day, and the importance of the forecast factors increases in the forecast for tomorrow and the day after tomorrow. In addition, date, temperature, humidity, and atmospheric pressure all show high importance regardless of time and concentration. Based on the importance of these factors, the LRP_DNN prediction model is developed. As a result, the ACC(accuracy) and POD(probability of detection) are improved by up to 5%, and the FAR(false alarm rate) is improved by up to 9% compared to the previous DNN model.
Keywords
Layer-wise Relevance Propagation (LRP); Explainable Artificial Intelligence (XAI); $PM_{2.5}$ Forecasting Model; Deep Neural Network (DNN);
Citations & Related Records
연도 인용수 순위
  • Reference
1 E.H. Shortliffe and B.G. Buchanan, "A Model of Inexact Reasoning in Medicine," Mathematical Biosciences, Vol. 23, Issues 3-4, pp. 351-379, 1975.   DOI
2 G. Montavon, S. Lapuschkin, A. Binder, W. Samek, and K.R. Muller, "Explaining Nonlinear Classification Decisions with Deep Taylor Decomposition," Pattern Recognition, Vol. 65. pp. 211-222, 2017.   DOI
3 NIER, A Development of Short-term Prediction Tool for PM10 and PM2.5 Concentrations using Artificial Intelligence (I), 11-1480523-0003767-01: NIER-SP2018-289, 2019.
4 U.S. EPA, Guidance on the Use of Models and Other Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.5, and Regional Haze, EPA-454/B-07-002, 2007.
5 NIER, A Study of Construction of Air Quality Forecasting System Using Artificial Intelligence(I), 11-1480523-0003221-01: NIER-SP 2017-148, 2017.
6 G. Yang, H. Lee, and G. Lee, "A Hybrid Deep Learning Model to Forecast Particulate Matter Concentration Levels in Seoul, South Korea," Atmosphere, Vol. 11, Issue 4, No. 348, 2020.
7 T. Xayasouk, H. Lee, and G. Lee, "Air Pollution Prediction Using Long Short-Term Memory (LSTM) and Deep Autoencoder (DAE) Models," Sustainablility, Vol 12, Issue 6, No. 2570, 2020.
8 S. Yu and Y. Jeon, "Improvement of PM10 Forecasting Performance Using DNN and Secondary Data," Journal of Korea Multimedia Society, Vol. 22, No. 10, pp. 1187-1198, 2019.
9 S. Yu, "Development of PM10 Forecasting Model for Seoul Based on DNN Using East Asian Wide Area Data," Journal of Korea Multimedia Society, Vol. 22, No. 11, pp. 1300-1312, 2019.
10 B.S. Freeman, G. Taylor, B. Gharabaghi, and J. The, "Forecasting Air Quality Time Series Using Deep Learning," Journal of the Air & Waste Management Association, Vol. 68, No. 8, pp. 866-886, 2018.   DOI
11 F. Biancofiore, M. Busilacchio, M. Verdecchia, B. Tomassetti and E. Aruffo, et al, "Recursive Neural Network Model for Analysis and Forecast of PM10 and PM2.5," Atmospheric Pollution Research, Vol. 8, Issue. 4, pp. 652-659, 2017.   DOI
12 Y. Koo, H. Yun, H. Kwon, and S. Yu, "A Development of PM10 Forecasting System," Journal of Korean Society for Atmospheric Environment, Vol. 26, No. 6, pp. 666-682, 2010.   DOI
13 J. Fan, Q. Li, J. Hou, X. Feng, H. Karimian, and S. Lin, "A Spatiotemporal Prediction Framework for Air Pollution Based on Deep RNN," Proceeding of ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W2, 2017 2nd International Symposium on Spatiotemporal Computing, pp. 15-22, 2017.
14 W. Lu, W. Wang, X. Wang, S. Yan, and J.C. Lam, "Potential Assessment of A Neural Network Model with PCA/RBF Approach for Forecasting Pollutant Trends in Mong Kok Urban Air, Hong Kong," Environmental Research, Vol. 96, No. 1, pp. 79-87, 2004.   DOI
15 S. Yu, Y. Jeon, and H. Kwon, "Improvement of PM10 Forecasting Performance Using Membership Function and DNN," Journal of Korea Multimedia Society, Vol. 22, No. 9, pp. 1069-1079, 2019.   DOI
16 S. Park, M. Kim, M. Kim, H. Namgung, K. Kim, K. Cho, et al., "Predicting PM10 Concentration in Seoul Metropolitan Subway Stations Using Artificial Neural Network (ANN)," Journal of Hazardous Materials, Vol. 341, pp. 75-82, 2018.   DOI
17 S. Yu, Y. Koo, and H. Kwon, "Inverse Model Parameter Estimation Based on Sensitivity Analysis for Improvement of PM10 Forecasting," Journal of Korea Multimedia Society, Vol. 18, No. 7, pp. 886-894, 2015.   DOI
18 H. Bae, S. Yu, and H. Kwon, "Fast Data Assimilation Using Kernel Tridiagonal Sparse Matrix for Performance Improvement of Air Quality Forecasting," Journal of Korea Multimedia Society, Vol. 20, No. 2, pp. 363-370, 2017.   DOI
19 M.V. Lent, W. Fisher, and M. Mancuso, "An Explainable Artificial Intelligence System for Small-Unit Tactical Behavior," Proceedings of the 16th Conference on Innovative Applications of Artificial Intelligence, pp. 900-907, 2004.
20 M.D. Zeiler and R. Fergus, "Visualizing and Understanding Convolution Networks," Proceeding of 13th European Conference on Computer Vision, pp. 818-833, 2014.
21 Y. Jo, H. Lee, L. Chang, and C. Kim, "Sensitivity Study of the Initial Meteorological Fields on the PM10 Concentration Prediction Using CMAQ Modeling," Journal of Korean Society for Atmospheric Environment, Vol. 33, No. 6, pp. 554-569, 2017.   DOI
22 U.S. EPA, Guidelines for Developing an Air Quality(Ozone and PM2.5) Forecasting Program, EPA-456/R-03-002, 2003.