| Summary: | 碩士 === 逢甲大學 === 統計與精算所 === 94 === The present paper will make the material analysis using the non-linear mixed model. In the daily life, has many longitudinal data and repeat measure. In the traditional statistics processing method majority by the linear model, like ANOVA or time series method (ARIMA).
Nowadays the non-linear mixed model joins the non-linear model in the fixed effect and random effects. This kind of statistical model more and more receives the widespread application to the longitudinal data. The growth curve function is in the longitudinal data research major part. The emphasis describes individual by the time mechanism the growth situation. This research will point out the Taiwan several industries (1954~2005) will use the electric quantity, the use growth function to perform to describe. Using SAS nlin and the nlmixed procedure, makes the model and the parameter matches suitably with the estimate, tries to discover the appropriate model.
|
|---|
