Please use this identifier to cite or link to this item:
標題: Validity of Dynamic Models for Event-History Analysis
作者: 吳宏達
關鍵字: 數學類;基礎研究;dynamic modeling;動態模型;事件史;平賭過程;復發事件;產卵;適合度;模型有效性;event-history;martingale process;recurrent event;oviposition;goodness-of-fit;model validity
This project studies the problem with emerging importance and interests: thedynamic modeling. We focus on testing the dynamic model (or rule) through the constructionof martingale estimating equation process which offers the basis for parameter estimation andmodel goodness-of-fit testing. By this approach, dynamic rule is tested through a techniquethat statisticians are familiar with. Firstly, we will define the meaning of dynamics instatistical language. Examples of possible applications will then be given, including theanalysis of recurrent event data and of the bean weevil's oviposition process. Simulationresults and actual data analysis are to be presented towards building up a scheme for visualgoodness-of-fit films.

本計劃研究日趨重要且有趣之動態模型檢證問題,藉由平賭過程來建構估計方程式,並得到有效的參數估計與適合度檢定。這本是統計學家所熟知的方法與技巧。首先,我們會以統計語彙定義出動態模型之意義。接著,以兩個例子提供説明: 包括復發事件資料分析,以及豆象產卵行為過程之分析皆是。我們會給出電腦模擬研究的結果,並分析實際資料以佐證說明我們的方法。最後,我們亦將試圖以影片的方式,呈現動態模型的適合度問題。
其他識別: NSC97-2118-M005-005
Appears in Collections:應用數學系所

Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.