English  |  正體中文  |  简体中文  |  Items with full text/Total items : 49621/84835 (58%)
Visitors : 7689774      Online Users : 63
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/32889


    Title: 重複捕取實驗中有異質性測量誤差時的迴歸校正估計方法
    Other Titles: Regression calibration estimation in capture-recapture experiment with heterogeneous measurement errors
    Authors: 翁振瑋;Weng, Chen-wei
    Contributors: 淡江大學數學學系碩士班
    黃逸輝;Huang, Yih-huei
    Keywords: 重複捕取實驗;測量誤差;異質性;母體總數;迴歸校正方法;Capture-recapture experiment;measurement error;heterogeneous;population size;RRC(refined regression calibration)
    Date: 2008
    Issue Date: 2010-01-11 02:56:18 (UTC+8)
    Abstract: 重複捕取(capture-recapture)實驗是一個常用來估計母體總數的方法,對於不同的重複捕取實驗而言,統計上已經發展出許多相關的分析方法和推論。異質性模型(heterogeneous model)的概念便是其中之一,即是考慮不同個體的異質性,如身高、體重、性別或其他可能影響捕取機率的因子(Pollock(1974))。其中邏輯式迴歸模型可能用於描繪動物被捕取的機率與其自變數之間的關係。理想上,自變數可以在動物被捕取時準確被測量,但實際上,它往往含有隨機誤差。

    在過去的文獻中,假設測量誤差的變異數不論捕取次數多寡皆是一樣的,似乎蘊含著每隻動物只會被測量一次自變數,在常理上較少使用。由於並非所有觀測對象只會被捕取一次,有些會被捕取一次以上,被捕取多次的觀測值平均值當然也會比較精準,因此我們應將它們全部使用在我們的估計式中,在本文我們將使用RRC的概念利用觀測值平均值進行迴歸係數和母體總數的估計。最後以模擬方法考量自變數與隨機誤差在不同分配之下估計方法的表現,並計算其平均估計值、估計值的樣本標準差、平均估計標準差、樣本均方根差(RMSE)與95%信賴區間涵蓋率(CP),並對結果加以討論。
    Capture-recapture is a commonly used method in estimating the population size.Ideally, covariate can be measured accurately for the statistical analysis.However it is usually measured with measurement error in practice.Some existent method had been developed to accommodate the measurement error,but these methods assume the variance of measurement error is homogeneous.This assumption implies that for each captured animal,the investigator measured its covariate only once though it may be captured more than one time.

    It is quite possible in practice that each animal''s covariate is measured everytime whenever it is captured, and the average of the observed value is more accurate than a single observed value.Therefore we constructed an estimating method using the average value based on the concept of refined regression calibration.It turns out that our method is also suitable when the variance of measurement error is heterogeneous.Simulation results are conducted for different kind of variance of measurement error and three kind of distribution of covariates.
    Appears in Collections:[數學學系暨研究所] 學位論文

    Files in This Item:

    File SizeFormat
    0KbUnknown195View/Open

    All items in 機構典藏 are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback