Assessing the Potential of Wild Mushrooms as Bioindicators for Environmental Pollution Prediction Using Machine Learning

Machine learning models, including predictive algorithms and neural networks, were used to compare data and make predictions regarding the effect of pollution on fungi based on soil and air data.Wild mushrooms were selected due to their high capacity for heavy metal uptake and ecological sensitivity, making them effective indicators of environmental pollution.Soil and fungal sampling was performed at varying distances from pollution sources.The samples underwent various chemical analyses to determine metal content, with metal concentrations expressed in mg/kg.The bioaccumulation factor was calculated, and heavy metal concentrations were measured using Atomic Absorption Spectroscopy (AAS).Orange Data Mining was used to apply machine learning algorithms, specifically neural networks, to predict the effects of pollution on metal accumulation in fungi based on soil and air measurements.Machine learning forecasts further suggested that fungi located closer to polluted sites tend to accumulate heavy metals, with lead accumulating at 6 mg/kg and cadmium at 2.67 mg/kg.Neural network forecasts showed good consistency with the actual values of bioaccumulation to indicate the possibilities of the algorithm in forecasting rates of pollution of soils and air with high levels of efficiency, particularly the heavy metals deposited within the mushrooms.Wild mushrooms from polluted areas were found to have greater efficiency in the bioaccumulation of heavy metals due to their very high absorption capability.Machine learning algorithms also provided correct results in predicting the effects of pollution on the environment, signifying the effectiveness of wild mushrooms as ideal bioindicators.