Based on the data of urban air quality,the normal classified value and corresponding probabitity of air quality have been obtained by Box Cox transformation and choosing the representative points of the standard normal distribution.
On the basis of research on parameter estimation method of Box Cox transformation model, drawbacks of MLE is analyzed and a method of parameter estimation to Box Cox transformation model is proposed in this paper,which is a generalized method of simulated moment based on residuals. Moreover,corresponding algorithm to this approach is investigated. Meanwhile,problem of estimation to Box Cox model with serial correlation and heteroscedasticity is solved.
This paper studies approximate confidence regions for the parameters and parameter subsets of transformations from geometric point of view. The results of the confidence regions are well suited for all transformations such as Box Cox transformation and shifted power transformation and so on.
Single factor model which describes effects of moisture, heated air temperature and air velocity on maize breakage susceptibility as well as soybean breakage susceptibility has been made by Box Cox conversion. Simulation of maize breakage susceptibility, together with soybean breakage susceptibility, could be done on the basis of the models.
In this paper. a class of nonlinear model一Box-Cox transformation model is considered. Theresult of its forecasting analysis indicates that forecasts based on the simple inverse transformation are biasedand should be corrected.
Contents of 11 heavy metals(Cu,Pb,Zn,Cd,Hg,Co,Ni,As,Cr,Mn and Fe)inthe sediments of the Changjiang River system were used for background contents calcula-tion. About 260 locations were selected,for sampling of both raw sediments and the finegrained parts(<63 μm)。 The frequency distribution and coeffieients of variation(Cv)werecalculated。
In the paper, we introduce the generalized BoxCox transformations, that is, we use the BoxCox transformations for dependent variables and some transformations for independent variables. And then, the transformation parameter is chosen by the maximum likelihood method. At last, a simple example is solved by using the transformation.