Forest Resource Model Heteroscedasticity

Carries on the performance forecast to the forest resources is a new problem of forest ecology construction research.On the one hand, few research is given by domestic and foreign scholars. According to the current forest condition,the forest resources plan and the ecology construction all will be extremely important for the forecast future development.Specially,the sustainable development long-term planning to implement the effective control and the plan construction to the current forest is the key point according to the forest ecology construction.However,the forest has the spatial distribution to be broad,has the long growth cycle,has multitudinous affects factors,has the complex structure,has the comprehensive function;has the limited memory and the limited relevance,renewability;The trend of development will have the uncertainty in the future;The forest also has the natural law and the economic law interweave mutually and so on;The forest was a disperser structure random dynamic large-scale system. According to the forest resources characteristic,we may use the random time-variable dynamic model to study,these models have the special characteristic of dimension high,multi-objectives and structure.In the classics regression analysis,the observed value variance homogeneity is a very basic hypothesis. Under this hypothesis,we are able to carry on conventional the statistical inference.If variance non-homogeneity moreover unknown,then the regression analysis will meet many questions. But,some examples indicated that,in the return model,the variance homogeneity supposition rationality is dubious truly,therefore the variance homogeneity examination is extremely essential. This article mainly study the different variance examination question of the single input list output forest resources model and the multi-input list output model,and obtained a sequence good result.This article main research work summarizes as follows:First,in the second chapter,I use the application variance structurization method to study the variance homogeneity examination question of the single input list output forest resources model. The present,I obtained variance homogeneity examination likelihood ratio test statistics and the score test statistics of the forest resources single input list output model.Next,I use the parameter orthogonalization method to obtain the variance homogeneity examination revision likelihood ratio test statistics of the single input list output forest resources model and the revision score test statistics. Finally,I has given each kind of test statistics value simulation,and carries on the example analysis.Finally indicated:The revision likelihood ratio examination surpasses the likelihood ratio examination;The revision score examination surpasses the score examination.Second,in the third chapter,I use the application variance structurization method to study the variance homogeneity examination question of the multi-input list output forest resources model. The present,I obtained variance homogeneity examination likelihood ratio test statistics and the score test statistics of the forest resources multi-input list output model.Next,I use the parameter orthogonalization method to obtain the variance homogeneity examination revision likelihood ratio test statistics of the multi-input list output forest resources model and the revision score test statistics. Finally,I has given each kind of test statistics value simulation,and carries on the example analysis.Finally indicated:The modified likelihood test examination surpasses the likelihood examination;The modified score examination surpasses the score examination.In summary, this article systematically has comprehensively studied the forest resources single input list output model and the multi-input list output model variance homogeneity examination question, obtained a series of test statistics. This article uses the method is succinct, effective; Obtains the result is correct.This article has provided one new method for the forest resources research, has provided the correct model for the actual forecast.This article has the important guiding sense to the forest resources practical work.