Open AccessWolf is Coming—Dynamic Classification Prediction Model of Vespa Mandarinia
Author(s) -
Yuexiang Yang,
Haomiao Niu,
Jiao Liu
Publication year - 2021
Publication title -
journal of environmental and earth sciences
Language(s) - English
Resource type - Journals
ISSN - 2661-3190
DOI - 10.30564/jees.v3i1.2924
Subject(s) - artificial neural network , computer science , support vector machine , artificial intelligence , data set , boundary (topology) , interpolation (computer graphics) , pattern recognition (psychology) , data mining , image (mathematics) , mathematics , mathematical analysis
Given the threat of Vespa mandarinia invasion to ecological balance, according to the data and information provided, the dynamic reproduction model of Vespa mandarinia is established by using natural domain interpolation, and the variation law of total bumblebee with time, latitude, and longitude is obtained. At the same time, we established the classification prediction model by using a neural network and established the mapping relationship between time and space to evaluation grade.We meshed the area provided by the title, assigned values to the location of Vespa mandarinia (VM), and established a VM diffusion model with natural neighborhood interpolation. Its propagation process is simulated by cellular automata. It is determined that VM spreads in a circular shape centered at (122.93174°W, 48.93457°N) and (122.57376°W, 49.07848°N) in the Washington area, with the farthest distance being 1184.4 km and 985 km respectively.We set up a classification prediction model for better classification. According to the image upload time and location, SVM and neural network are used for classification prediction, and the classification accuracy is 74.26% and 97.60%, respectively, and the neural network has higher classification accuracy. So we choose the neural network.