Identifying Effects of Air Pollutants in California Bee Population: A Machine Learning Case Study

This case study examined the role of air quality in bee population decline and determined that accurate models can be built using machine learning to predict bee population based on presence of pollutants in the air.  The findings of this project indicated that the k-NN model was the most successful model in predicting honeybee population using CO, NO2, ozone, PM10, PM2.5, year, and county population density as variables.  Ozone and PM10 concentrations were found to be the most important drivers of honeybee population decline, of all variables considered.  The results of this project affirmed the significance of the negative effects of climate change and air pollution on both vulnerable and critically important honeybee populations.  These findings were used to inform recommendations for managing honeybee populations in California and other important agricultural states and provide direction for future research opportunities in this field.