Application of GA-SVM for County Ecological Security Prediction of Land Resources

Abstract

Ecological security prediction of land resources play an important role in sustainable utilization of land resources and improve benefit of healthy development of urbanization in China. So far, many methods for regional ecological security prediction have been proposed. According to the county level of ecological security prediction of land resources data which is large scale and imbalance, this paper presented a support vector machine (SVM) model to predict the county level of ecological security of land resources. However, the performance strongly depends on the right selection of the parameters of the SVM model. In order to improve the discrimination precision of SVM in prediction, a Genetic Algorithm (GA) was used to optimize SVM parameters in the solution space. We selected Guanzhong urban agglomeration, a typical urban agglomeration of west China, as a case. The method was compared with SVM, artificial neural network, decision tree, logistic regression and naive Bayesian classifier regarding the county level of ecological security of land resources prediction. The result shows that the improved SVM was much better than other models on accuracy rate, hit rate, covering rate and lift coefficient. GA-SVM model is a potential effective candidate for the prediction of ecological security of land resources.