Breeding model of Bumblebee based on linear regression and BP neural network

Bumblebee is the natural enemy of many bees. If the reproduction of bumblebee is not predicted, it may destroy the balance of the whole colony. Firstly, Positive ID was selected by preprocessing the datasets and dividing the resulting datasets according to laboratory status. In this paper, linear regression is used to analyze the data set of time series, and BP neural network algorithm is further used to analyze the data set of the relationship witness report and detection date of the longitude and latitude bumblebee, and finally the prediction and analysis are carried out. The experimental results show that the transmission of bumblebees can be predicted by the longitude, latitude and detection date reported by witnesses. The establishment of the model is helpful for ecological researchers to predict the overall reproduction of bumblebees in space and provide support for the protection of bee colonies