Forecasting Sudden Drops Of Temperature In Pre-Overwintering Honeybee Colonies

Bees are the main pollinators of most cultivated and wild plant species. Unfortunately, due to the pesticide use, habitat loss, and climate change, they are declining worldwide. A machine learning model to anticipate temperature drops in honeybee colonies. The long short-term memory (LSTM) algorithm was applied to five real datasets with the following input factors: (i) internal temperature, (ii) internal humidity, (iii) mean fanning, (iv) mean noise, (v) mass, and (vi) external temperature. The results showed that the proposed machine learning model could predict the temperature 24 h in advance with an RMSE error of only 0.5%. (c) 2021 IAgrE. Published by Elsevier Ltd. All rights reserved.