Application of Coupling Model of Projection Pursuit Partial Least-Square Regression Based on Real Coded Accelerating Genetic Algorithm in Land Use Change Forecasting
In study of Land Use Change forecasting, lots of methods have been developed ,such as Markov model、BP algorithm、Canonical correlation analysis, least-squares regression analysis ,but these methods have deficiency in decision and often inadequate in sample size. In response to these deficiencies，projection pursuit Partial Least-Square Regression based on real coded accelerating genetic algorithm model is developed to analyze and predict land use change in Yanji City. The computation results show that the relative error of Coupling Model of Partial Least-Square Regression Based on Projection Pursuit is smaller than traditional Partial Least-Square Regression model’s, and it has improved the prediction precision evidently.
Weiguo Pan, Jianxing Ren and Yongguang Li
H. T. Wang et al., "Application of Coupling Model of Projection Pursuit Partial Least-Square Regression Based on Real Coded Accelerating Genetic Algorithm in Land Use Change Forecasting", Advanced Materials Research, Vols. 347-353, pp. 1774-1777, 2012