具冪次轉換，隨機效應及AR(g)相關之成長曲線模型的參數估計與預測

Abstract

在本篇論文中，我們著眼於以貝氏方法及最大概似法兩種 觀點來分析㆒具有幂次轉換，隨機效應及 AR(g)相關之廣義成 長曲線模型。所含蓋的統計推論除了參數的估計之外，尚包括 有對未觀測值的預測。另外，實際資料及模擬資料的分析比較 在文中均有詳細的討論。

In this paper, we devote ourselves to a generalized growth curve model with power transformation, random effects and AR($g$) dependence via Bayesian and Maximum-Likelihood (ML) approaches. Inferences on the parameters as well as the future value are discussed. Some numerical results with real and simulated data are also given.\\ {\bf Key words}: Approximation, Bayesian, Longitudinal data, Maximum likelihood, Markov Chain Monte Carlo, Real data, Simulated data.