Abstract
An increasing number of chemicals found in the environment potentially pose a threat to organisms such as fish. Models for risk assessment are vital resources that enable possible measurements of the hazards associated with chemical exposure. Traditional monitoring techniques and experimental procedures, however, are unable to keep up with the compounds that are becoming more and more implicated in environmental problems. Furthermore, a significant amount of data invariably results in inaccuracies. Here, we proposed an integrated approach that combines machine learning and fuzzy logic mathematical methods, assessing the risks associated with chemical exposure from contaminated fish with the least amount of data entry possible. We predicted the concentrations of organic contaminants in the environment, serving as a baseline for quantifying the fuzzy risks during household thermal processing of the fish. With a mean R2 value of 0.78, concentration of chemicals in the aquatic environment emerged as the most influential factor in predicting bioconcentration factors. Heptachlor, Endosulfan-sulfate, Endrin, and Endrin aldehyde are four high-risk pesticides throughout the entire processing process compared to others. The findings underscore the importance of considering processing methods and environmental factors in order to ensure food safety.