Explaining deep learning models for ozone pollution prediction via embedded feature selection

被引:12
作者
Jimenez-Navarro, M. J. [1 ]
Martinez-Ballesteros, M. [1 ]
Martinez-Aalvarez, F. [2 ]
Asencio-Cortes, G. [2 ]
机构
[1] Univ Seville, Dept Comp Languages & Syst, ES-41012 Seville, Spain
[2] Pablo de Olavide Univ, Data Sci & Big Data Lab, ES-41013 Seville, Spain
关键词
Pollution; Time series forecasting; Ozone; Feature selection; Deep learning; XAI; TIME;
D O I
10.1016/j.asoc.2024.111504
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Ambient air pollution is a pervasive global issue that poses significant health risks. Among pollutants, ozone ( O 3 ) is responsible for an estimated 1 to 1.2 million premature deaths yearly. Furthermore, O 3 adversely affects climate warming, crop productivity, and more. Its formation occurs when nitrogen oxides and volatile organic compounds react with short -wavelength solar radiation. Consequently, urban areas with high traffic volume and elevated temperatures are particularly prone to elevated O 3 levels, which pose a significant health risk to their inhabitants. In response to this problem, many countries have developed web and mobile applications that provide real-time air pollution information using sensor data. However, while these applications offer valuable insight into current pollution levels, predicting future pollutant behavior is crucial for effective planning and mitigation strategies. Therefore, our main objectives are to develop accurate and efficient prediction models and identify the key factors that influence O 3 levels. We adopt a time series forecasting approach to address these objectives, which allows us to analyze and predict future O 3 behavior. Additionally, we tackle the feature selection problem to identify the most relevant features and periods that contribute to prediction accuracy by introducing a novel method called the Time Selection Layer in Deep Learning models, which significantly improves model performance, reduces complexity, and enhances interpretability. Our study focuses on data collected from five representative areas in Seville, Cordova, and Jaen provinces in Spain, using multiple sensors to capture comprehensive pollution data. We compare the performance of three models: Lasso, Decision Tree, and Deep Learning with and without incorporating the Time Selection Layer. Our results demonstrate that including the Time Selection Layer significantly enhances the effectiveness and interpretability of Deep Learning models, achieving an average effectiveness improvement of 9% across all monitored areas.
引用
收藏
页数:11
相关论文
共 46 条
[1]   Generation forecasting employing Deep Recurrent Neural Network with metaheruistic feature selection methodology for Renewable energy power plants [J].
Alshammari, Abdulaziz .
SUSTAINABLE ENERGY TECHNOLOGIES AND ASSESSMENTS, 2023, 55
[2]  
[Anonymous], 2019, Healthy Environments for Healthier Populations: why Do They Matter, and What Can We Do?
[3]  
Benavoli A, 2017, J MACH LEARN RES, V18
[4]  
Bhaskar Kura S V., 2013, Computational Water, Energy, and Environmental Engineering, V2, P1, DOI DOI 10.4236/CWEEE.2013.22B001
[5]  
Cancela B., 2020, Clin. Orthop. Related Res.
[6]   Analysis and improvements on feature selection methods based on artificial neural network weights [J].
da Costa, Nattane Luiza ;
de Lima, Marcio Dias ;
Barbosa, Rommel .
APPLIED SOFT COMPUTING, 2022, 127
[7]   A comprehensive survey on feature selection in the various fields of machine learning [J].
Dhal, Pradip ;
Azad, Chandrashekhar .
APPLIED INTELLIGENCE, 2022, 52 (04) :4543-4581
[8]   Accelerating Deep Neural Networks implementation: A survey [J].
Dhouibi, Meriam ;
Ben Salem, Ahmed Karim ;
Saidi, Afef ;
Ben Saoud, Slim .
IET COMPUTERS AND DIGITAL TECHNIQUES, 2021, 15 (02) :79-96
[9]  
European Parliament and of the Council on ambient air quality and cleaner air for Europe, 2008, Directive2008/50/EC, 2008.
[10]   A novel approach to forecast urban surface-level ozone considering heterogeneous locations and limited information [J].
Gomez-Losada, Alvaro ;
Asencio-Cortes, G. ;
Martinez-Alvarez, F. ;
Riquelme, J. C. .
ENVIRONMENTAL MODELLING & SOFTWARE, 2018, 110 :52-61