Study of plant communities based on supervised learning model SVM

Abstract

Numerous studies have shown that the number of different species plays an important role in how the plant community adapts during successive multi-generational drought cycles. In this paper, we first collected relevant data and completed the pre-processing of the data. Then the community ecological environment index was quantitatively analyzed through indicators such as biodiversity and birth mortality in the region, and the correlation between the characteristic X indicators was obtained through correlation analysis, and the supervised learning relationship between the characteristic indicators, indicators and ecological environment index was established, and the best SVM model was obtained through Bayesian optimization and network parameter synchronization. The final model accuracy is 91.6%, and the root mean square error is less than 0.382, which is adaptable and strong scientific. Based on this model, the number of species in the index, the correlation between species and the value of the environmental quality index kept changing, the data changed in the model to the corresponding changes in the ecological environmental index, in order to see the impact of different environments and species in the ecosystem, a qualitative analysis was performed, and finally the impact of the number and type of species on the plant community was studied.