Revealing Toluene Behaviour in the Atmosphere Based on Coupling of Metaheuristics, Xgboost, and Shap


DOI: https://doi.org/10.15308/Sinteza-2023-17-22
Authors:
Gordana Jovanović
Mirjana Perišić
Svetlana Stanišić
Nebojša Bačanin Džakula
Andreja Stojić

Download full paper

Keywords:
Machine Learning, Extreme Gradient Boosting, Metaheuristics, Explainable Artificial Intelligence, Volatile Organic Compounds


Abstract:
This study used an improved version of the reptilian search algorithm to investigate atmospheric patterns of toluene and its interactions with other polluting species under different environmental conditions. Toluene is a harmful aromatic hydrocarbon known for its role in the formation of secondary atmospheric pollutants. In this study, a two-year database of hourly pollutant concentrations, such as toluene, was analysed. The results were validated against other models using metaheuristic algorithms, and Shapley's additive explanations method was used to interpret them. The findings indicated a distinct correlation between toluene and m,p-xylene, and the study described the environmental conditions that influence their interactions. Overall, this research highlights the significance of using advanced analytical techniques to better understand the relationships between pollutants and their behaviour in different environmental conditions.

CITATION:

IEEE format

G. Jovanović, M. Perišić, S. Stanišić, N. Bačanin Džakula, A. Stojić, “Revealing Toluene Behaviour in the Atmosphere Based on Coupling of Metaheuristics, Xgboost, and Shap,” in Sinteza 2023 - International Scientific Conference on Information Technology, Computer Science, and Data Science, Belgrade, Singidunum University, Serbia, 2023, pp. 17-22. doi:10.15308/Sinteza-2023-17-22

APA format

Jovanović, G., Perišić, M., Stanišić, S., Bačanin Džakula, N., Stojić, A. (2023). Revealing Toluene Behaviour in the Atmosphere Based on Coupling of Metaheuristics, Xgboost, and Shap. Paper presented at Sinteza 2023 - International Scientific Conference on Information Technology, Computer Science, and Data Science. doi:10.15308/Sinteza-2023-17-22

BibTeX format
Download

RefWorks Tagged format
Download