Author(s)

Chubing Chen, Nanyu Zheng and Minghua Huang

Corresponding Author

Chubing Chen

ABSTRACT

The Asian hornet is the largest wasp species in the world and is a greater potential threat to native bee populations, and the occurrence of invasive Asian hornets in Washington State poses a significant challenge to public safety and native species. First, we pre-processed the data for better analysis, including data quantification and balancing the data set using the SMOTE algorithm. And two models were built. One is an AGH propagation prediction model based on LSTM algorithm. The other is a prediction error classification probability model based on BP neural network. Second, we applied model 1 to predict the latitude and longitude of the new nest of Asian hornet as [48.88721431, -122.47042182], and the mean square error of the model was 0.0017433 and the distance from the previous nest was 29. 234 km, which was consistent with the nesting habit of the queen bee. Third, we used the four dimensions of Detection Date, Notes, Latitude, and Longitude as the input parameters of Model 2 for training, and obtained a goodness-of-fit of 0. 7862.The unprocessed dataset was predicted using model 2 to obtain the priority investigation report ranking of the Unprocessed dataset, where the sighting report with Global ID 26DF8E2 -DAOC- 4F87-A65A-2331 15BAFCCD was ranked the highest and should be investigated most priority.

KEYWORDS

LSTM model, BP neural network, genetic algorithm