{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T04:30:33Z","timestamp":1787027433071,"version":"3.56.0"},"reference-count":59,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,8,4]],"date-time":"2025-08-04T00:00:00Z","timestamp":1754265600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>\n                    The study addresses the problem of nonlinear characteristics of common air pollutants by proposing a deep learning time-series model based on the long short-term memory (LSTM) integrated with a generalized additive model (GAM). LSTM model captures both nonlinear relationships and temporal long-term dependencies in time-series data, and GAM provides insight into the statistical relationship between selected features and the target pollutant. The post-hoc eXplainable artificial intelligence (xAI) technique, local interpretable model-agnostic explanation (LIME), further explains the nonlinearity. Finally, causal inference was determined on the impact of the air pollutants relationship, thereby offering further interpretability in which deep learning models are deficient. Meteorological and air pollutant statistical records were leveraged from a Hantam (Karoo) air monitoring station in South Africa, and through a random sampling approach, synthetic data were generated for the city of Kimberley. The model was evaluated with the mean squared error (MSE), root mean squared error (RMSE) and mean absolute error (MAE) for different time-steps. The proposed referred to as long short-term memory generalized additive model based post-hoc eXplainable Artificial Intelligence (LSTM-GAM_xAI) model with a 10-day time-step and 5-day time-step for multiple pollutants prediction guaranteed least MSE. Though the causal effect analysis show no\n                    <jats:italic>p<\/jats:italic>\n                    -values (&amp;gt;0.88) for variables, the experiment results show that LSTM-GAM-xAI guaranteed the lowest MSE values across different time-steps.\n                  <\/jats:p>","DOI":"10.3389\/frai.2025.1620019","type":"journal-article","created":{"date-parts":[[2025,8,4]],"date-time":"2025-08-04T06:13:13Z","timestamp":1754287993000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":3,"title":["A hybrid long short-term memory with generalized additive model and post-hoc explainable artificial intelligence with causal inference for air pollutants prediction in Kimberley, South Africa"],"prefix":"10.3389","volume":"8","author":[{"given":"Israel Edem","family":"Agbehadji","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ibidun Christiana","family":"Obagbuwa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,8,4]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"345","DOI":"10.3390\/atmos13020345","article-title":"Ambient particulate air pollution and daily hospital admissions in 31 cities in Poland","volume":"13","author":"Adamkiewicz","year":"2022","journal-title":"Atmos."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"1352","DOI":"10.3390\/atmos15111352","article-title":"Systematic review of machine learning and deep learning techniques for spatiotemporal air quality prediction","volume":"15","author":"Agbehadji","year":"2024","journal-title":"Atmosphere, MDPI"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"1","DOI":"10.3390\/forecast7020015","article-title":"Mode decomposition bi-directional long short-term memory (BiLSTM) attention mechanism and transformer (AMT) model for ozone (O3) prediction in Johannesburg, South Africa","volume":"7","author":"Agbehadji","year":"","journal-title":"Forecasting"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"523","DOI":"10.3390\/atmos16050523","article-title":"Integration of explainable artificial intelligence into hybrid long short-term memory and adaptive Kalman filter for sulfur dioxide (SO2) prediction in Kimberley, South Africa","volume":"16","author":"Agbehadji","year":"","journal-title":"Atmos."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"111269","DOI":"10.1016\/j.envres.2021.111269","article-title":"Impact of ambient air quality standards revision on the exposure-response of air pollution in Tianjin, China","volume":"198","author":"Bai","year":"2021","journal-title":"Environ. Res."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"119241","DOI":"10.1016\/j.envres.2024.119241","article-title":"A multi-resolution ensemble model of three decision-tree-based algorithms to predict daily NO2 concentration in France 2005\u20132022","volume":"257","author":"Barbalat","year":"2024","journal-title":"Environ. Res."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"100","DOI":"10.17159\/caj\/2024\/34\/2.20100","article-title":"Ambient PM2.5, soot, black carbon and organic carbon levels in Kimberley, South Africa","volume":"34","author":"Becker","year":"2024","journal-title":"Clean Air J"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"120677","DOI":"10.1016\/j.atmosenv.2024.120677","article-title":"Prediction of urban ultrafine particle emission fluxes using generalized additive models","volume":"334","author":"Bitz","year":"2024","journal-title":"Atmos. Environ."},{"key":"ref9","volume-title":"NODE-GAM: neural generalized additive model for interpretable deep learning","author":"Chang","year":"2022"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"20","DOI":"10.2139\/ssrn.4232987","article-title":"An environmental disaster: a critical review of kimberlite diamond Mining in Kimberley, South Africa","author":"Charumbira","year":"2022","journal-title":"Int J Disaster Risk Reduct"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"102024","DOI":"10.1016\/j.jocs.2023.102024","article-title":"Ensemble Kalman filter for GAN-ConvLSTM based long lead-time forecasting","volume":"69","author":"Cheng","year":"2023","journal-title":"J Comput Sci"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"118437","DOI":"10.1016\/j.atmosenv.2021.118437","article-title":"Influence of weather and air pollution on concentration change of PM2.5 using a generalized additive model and gradient boosting machine","volume":"255","author":"Cheng","year":"2021","journal-title":"Atmos. Environ."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"3645","DOI":"10.1007\/s00180-024-01593-z","article-title":"Gradient-based smoothing parameter estimation for neural P-splines","volume":"40","author":"Dammann","year":"2025","journal-title":"Comput. Stat."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"100806","DOI":"10.1016\/j.spasta.2023.100806","article-title":"A generalized additive model (GAM) approach to principal component analysis of geographic data","volume":"59","author":"de As\u00eds L\u00f3pez","year":"2024","journal-title":"Spatial Statistics"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"107656","DOI":"10.1016\/j.atmosres.2024.107656","article-title":"Interpretable machine learning tools to analyze PM2.5 sensor network data so as to quantify local source impacts and long-range transport","volume":"311","author":"de Foy","year":"2024","journal-title":"Atmos. Res."},{"key":"ref16","first-page":"1","volume-title":"National air quality indicator - monthly data report for the northern Cape Province","year":"2019"},{"key":"ref17","author":"Dillon","year":"2025"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"8783","DOI":"10.1007\/s00521-019-04282-x","article-title":"A real-time hourly ozone prediction system using deep convolutional neural network","volume":"32","author":"Eslami","year":"2020","journal-title":"Neural Comput. & Applic."},{"key":"ref19","doi-asserted-by":"publisher","first-page":"1373","DOI":"10.1007\/s11869-024-01513-6","article-title":"Association between outdoor air pollutants and risk of acute exacerbation of chronic obstructive pulmonary disease in Xi\u2019an, China","volume":"17","author":"Fan","year":"2024","journal-title":"Air Qual. Atmos. Health"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"35790","DOI":"10.1007\/s11356-021-13268-z","article-title":"Short-term nitrogen dioxide exposure is associated with the spread of S. pyogenes-induced vulvovaginitis in prepubertal girls in Hangzhou, China","volume":"28","author":"Fang","year":"2021","journal-title":"Environ. Sci. Pollut. Res."},{"key":"ref21","first-page":"1","volume-title":"National air Quality Indicator - monthly data report for the northern Cape Province","year":"2024"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"27","DOI":"10.3390\/atmos14010027","article-title":"Hourly Valley concentration of air pollutants associated with increased acute myocardial infarction hospital admissions in Beijing, China","volume":"14","author":"Fu","year":"2023","journal-title":"Atmos."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"1018","DOI":"10.3390\/atmos11091018","article-title":"Assessment of the impact of PM2.5 exposure on the daily mortality of circulatory system in Shijiazhuang, China","volume":"11","author":"Fu","year":"2020","journal-title":"Atmos."},{"key":"ref24","doi-asserted-by":"publisher","first-page":"125642","DOI":"10.1016\/j.envpol.2025.125642","article-title":"Health effects associated with ozone in China: a systematic review","volume":"367","author":"Gao","year":"2025","journal-title":"Environ. Pollut."},{"key":"ref25","first-page":"39","volume-title":"Learning functional causal models with generative neural networks. Xplainable and interpretable models in computer vision and machine learning","author":"Goudet","year":"2018"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"136896","DOI":"10.1016\/j.scitotenv.2020.136896","article-title":"Ambient fine particulate matter and hospital admissions for ischemic and hemorrhagic strokes and transient ischemic attack in 248 Chinese cities","volume":"715","author":"Gu","year":"2020","journal-title":"Sci. Total Environ."},{"key":"ref27","doi-asserted-by":"publisher","first-page":"1164","DOI":"10.1007\/s12517-020-06179-9","article-title":"Assessment of ground-level ozone pollution with monitoring and modelling approaches in Makkah, Saudi Arabia","volume":"13","author":"Habeebullah","year":"2020","journal-title":"Arab. J. Geosci."},{"key":"ref28","doi-asserted-by":"publisher","first-page":"55","DOI":"10.1007\/s11869-020-00913-8","article-title":"Air quality pollutants and their relationship with meteorological variables in four suburbs of greater Sydney, Australia","volume":"14","author":"Haddad","year":"2021","journal-title":"Air Qual. Atmos. Health"},{"key":"ref29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.17159\/caj\/2021\/31\/2.8880","article-title":"Modeling tropospheric ozone and particulate matter in Tunis, Tunisia using generalized additive model","volume":"31","author":"Hammouda","year":"2021","journal-title":"Clean Air J"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"101763","DOI":"10.1016\/j.apr.2023.101763","article-title":"Predicting fine-scale daily NO2 over Mexico city using an ensemble modeling approach. Atmospheric","volume":"14","author":"He","year":"2023","journal-title":"Pollut. Res."},{"key":"ref31","doi-asserted-by":"publisher","first-page":"102463","DOI":"10.1016\/j.apr.2025.102463","article-title":"Estimating and characterizing spatiotemporal distributions of elemental PM2.5 using an ensemble machine learning approach in Taiwan. Atmospheric","volume":"16","author":"Huang","year":"2025","journal-title":"Pollut. Res."},{"key":"ref32","doi-asserted-by":"publisher","first-page":"5936","DOI":"10.3390\/su16145936","article-title":"Determination of optimal batch size of deep learning models with time series data","volume":"16","author":"Hwang","year":"2024","journal-title":"Sustainability"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"120307","DOI":"10.1016\/j.eswa.2023.120307","article-title":"Neural additive time-series models: explainable deep learning for multivariate time-series prediction","volume":"228","author":"Jo","year":"2023","journal-title":"Expert Syst. Appl."},{"key":"ref34","first-page":"1","article-title":"Interpreting interpretability: understanding data scientists' use of interpretability tools for machine learning","volume-title":"Proceedings of the 2020 CHI conference on human factors in computing systems","author":"Kaur","year":"2020"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s12599-024-00922-2","article-title":"Challenging the performance-interpretability trade-off: an evaluation of interpretable machine learning models","author":"Kruschel","year":"2025","journal-title":"Bus. Inf. Syst. Eng."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"ref37","doi-asserted-by":"publisher","first-page":"54","DOI":"10.1016\/j.jes.2024.01.057","article-title":"Meteorological and traffic effects on air pollutants using Bayesian networks and deep learning","volume":"152","author":"Lin","year":"2025","journal-title":"J. Environ. Sci. (China)"},{"key":"ref38","first-page":"150","article-title":"Intelligible models for classification and regression","volume-title":"Proceedings of the 18th ACM SIGKDD international conference on knowledge discovery and data mining","author":"Lou","year":"2012"},{"key":"ref39","doi-asserted-by":"publisher","first-page":"114073","DOI":"10.1109\/ACCESS.2024.3443810","article-title":"Two-stage feature engineering to predict air pollutants in urban areas","volume":"12","author":"Naz","year":"2024","journal-title":"IEEE Access"},{"key":"ref40","doi-asserted-by":"crossref","first-page":"121","DOI":"10.1016\/B978-0-12-374765-5.00007-3","article-title":"Chapter 7 basic algorithms for data mining: a brief overview","volume-title":"Handbook of statistical analysis and data mining applications","author":"Nisbet","year":"2009"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1016\/j.procs.2022.12.214","article-title":"Improving boosted generalized additive models with random forests: a zoo visitor case study for smart tourism","volume":"217","author":"Obster","year":"2022","journal-title":"Procedia Comput. Sci."},{"key":"ref42","doi-asserted-by":"publisher","first-page":"6","DOI":"10.1007\/s11222-023-10320-5","article-title":"Explainable generalized additive neural networks with independent neural network training","volume":"34","author":"Ortega-Fernandez","year":"2024","journal-title":"Stat. Comput."},{"key":"ref43","doi-asserted-by":"publisher","first-page":"104987","DOI":"10.1016\/j.envint.2019.104987","article-title":"Generalized additive models: building evidence of air pollution, climate change and human health","volume":"132","author":"Ravindra","year":"2019","journal-title":"Environ. Int."},{"key":"ref44","doi-asserted-by":"crossref","DOI":"10.1109\/CITSM56380.2022.9936037","article-title":"Prophet model performance analysis for Jakarta air quality forecasting","volume-title":"2022 10th international conference on cyber and IT service management (CITSM)","author":"Setianingrum","year":"2022"},{"key":"ref45","doi-asserted-by":"publisher","first-page":"3060","DOI":"10.1007\/s13198-024-02315-w","article-title":"Comparative analysis of machine learning techniques in air quality index (AQI) prediction in smart cities","volume":"15","author":"Sharma","year":"2024","journal-title":"Int. J. Syst. Assur. Eng. Manag."},{"key":"ref46","doi-asserted-by":"publisher","first-page":"1929","DOI":"10.5555\/2627435.2670313","article-title":"Dropout: a simple way to prevent neural networks from overfitting","volume":"15","author":"Srivastava","year":"2014","journal-title":"J Mach Learn Res"},{"key":"ref47","doi-asserted-by":"crossref","DOI":"10.1109\/ICCCNT56998.2023.10306465","article-title":"Air pollution monitoring & prediction system for Ahmedabad region using IOT and ML","volume-title":"2023 14th International Conference on Computing Communication and Networking Technologies (ICCCNT)","author":"Talati","year":"2023"},{"key":"ref48","doi-asserted-by":"crossref","DOI":"10.1109\/IJCNN.2011.6033450","article-title":"Forecasting tropospheric ozone concentrations with adaptive neural networks","volume-title":"The 2011 International Joint Conference on Neural Networks","author":"Taormina","year":"2011"},{"key":"ref49","doi-asserted-by":"publisher","first-page":"797","DOI":"10.1038\/s41588-025-02124-2","article-title":"Causal machine learning for single-cell genomics","volume":"57","author":"Tejada-Lapuerta","year":"2025","journal-title":"Nat. Genet."},{"key":"ref50","doi-asserted-by":"publisher","first-page":"1287","DOI":"10.1631\/FITEE.2200540","article-title":"LDformer: a parallel neural network model for long-term power forecasting","volume":"24","author":"Tian","year":"2023","journal-title":"Front Inform Technol Electron Eng."},{"key":"ref51","doi-asserted-by":"publisher","first-page":"2237","DOI":"10.5194\/amt-16-2237-2023","article-title":"Estimation of NO2 emission strengths over Riyadh and Madrid from space from a combination of wind-assigned anomalies and a machine learning technique","volume":"16","author":"Tu","year":"2023","journal-title":"Atmos. Meas. Tech."},{"key":"ref52","doi-asserted-by":"publisher","first-page":"94","DOI":"10.1007\/s10462-023-10698-8","article-title":"A review of predictive uncertainty estimation with machine learning","volume":"57","author":"Tyralis","year":"2024","journal-title":"Artif. Intell. Rev."},{"key":"ref53","doi-asserted-by":"publisher","first-page":"987","DOI":"10.3390\/atmos15080987","article-title":"A generalised additive model and deep learning method for cross-validating the North Atlantic oscillation index","volume":"15","author":"Wahiduzzaman","year":"2024","journal-title":"Atmos."},{"key":"ref54","doi-asserted-by":"crossref","DOI":"10.1109\/ACCESS51619.2021.9563281","article-title":"Multivariate air pollution levels forecasting","volume-title":"2021 2nd International Conference on Advances in Computing, Communication, Embedded and Secure Systems (ACCESS)","author":"Wattal","year":"2021"},{"key":"ref55","doi-asserted-by":"crossref","DOI":"10.1155\/2021\/9903518","volume-title":"Multivariate CNN-LSTM model for multiple parallel financial time-series prediction","author":"Widiputra","year":"2021"},{"key":"ref56","doi-asserted-by":"publisher","first-page":"120269","DOI":"10.1016\/j.atmosenv.2023.120269","article-title":"Improved O3 predictions in China by combining chemical transport model and multi-source data with machine learning techniques","volume":"318","author":"Xiong","year":"2024","journal-title":"Atmos. Environ."},{"key":"ref57","doi-asserted-by":"publisher","first-page":"102516","DOI":"10.1016\/j.jag.2021.102516","article-title":"Tree-based ensemble deep learning model for spatiotemporal surface ozone (O3) prediction and interpretation","volume":"103","author":"Zang","year":"2021","journal-title":"Int J Appl Earth Observat Geoinform"},{"key":"ref58","doi-asserted-by":"publisher","first-page":"107859","DOI":"10.1016\/j.jobe.2023.107859","article-title":"Application and enabling digital twin technologies in the operation and maintenance stage of the AEC industry: a literature review","volume":"80","author":"Zhang","year":"2023","journal-title":"J Build Eng"},{"key":"ref59","doi-asserted-by":"publisher","first-page":"1085","DOI":"10.1007\/s10651-024-00635-5","article-title":"A comparison of statistical and machine learning models for spatio-temporal prediction of ambient air pollutant concentrations in Scotland","volume":"31","author":"Zhu","year":"2024","journal-title":"Environ. Ecol. Stat."}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1620019\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,4]],"date-time":"2025-08-04T06:13:16Z","timestamp":1754287996000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1620019\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,4]]},"references-count":59,"alternative-id":["10.3389\/frai.2025.1620019"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1620019","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,4]]},"article-number":"1620019"}}