Predicting odor concentration for environmental sustainability: a comparison among Machine Learning methods

Odor emissions from a treatment plant is one of the major environmental issues that results in negative health consequences and repercussions on economic, commercial, and touristic activities. To address this problem an accurate assessment of the odor sources is of crucial interest. In this paper, different machine learning methods are applied to identify the most suitable model to estimate odor concentrations through the responses of a multiparametric system. It is observed that random forest regression method shows superior performance compared to the other methods. In this context, advanced data analytics technologies, such as machine learning methods, have provided data-driven decision-making capabilities to address the challenges that arise in the analysis and evaluation of a sustainable development. The findings of the proposed study can help implement proactive actions to minimize the effects of odors and prevent any potential health and environmental concerns.