臺灣博碩士論文加值系統

在自然與工業界中，生物與產品的壽命往往是研究的重點，例如在醫學的臨床實驗中，有的病人在實驗未結束之前，可能因為某些因素而無法繼續實驗，又如在實驗中可能有一些實驗者對實驗有危險性，或不適合實驗的要求，而要提前結束實驗，為了解決這些問題，所以發展出數種設限與逐步移除的方法。本論文是探討二項與隨機移除型二逐步設限下極值分配和柏拉圖分配資料的統計分析。
在文中，首先討論分配中參數的最大概似估計量與特性，並對隨機移除與二項移除兩方法分別探討，再推導極值分配與柏拉圖分配在隨機與二項移除的期望時間，在分配的參數作一些變化計算出期望時間，再來加以分析極值分配與柏拉圖分配在二項與隨機移除型二逐步設限下的期望時間。

Studies in the lifetimes of the organisms and products are often the main research topics in the nature and industries. The past research developed some censored methods. Before the experiments did not end yet, some patients might not proceed the experiments due to some factors. Furthermore, the experiments were ended early because some experimenters might be hazardous for the experiments during the experiments or not tally with the demands of the experiments. In order to resolve these problems, progressive removal was developed. This thesis mainly concentrates on the statistical analysis of Extreme-value distributed and Pareto distributed lifetime data under Type II progressive censoring with binomial and random removals.
To begin with, the thesis discusses the maximum likelihood estimates and characteristics of the parameters in the distributions. We respectively narrate binomial and random removals. Then, we separately deduce the expected values of Extreme-value distribution and Pareto distribution under binomial and random removals. Adjusting the parameter values of the distributions counts up to the expected time. Finally, we farther analyze the expected time of Extreme-value distribution and Pareto distribution under Type II progressive censoring with binomial and random removals.

目錄
表目錄………………………………………………………………………III
圖目錄………………………………………………………………………IV
第一章 緒論…………………………………………………………………1
1-1 前言………………………………………………………………………1
1-2研究背景與動機…………………………………………………………1
1-3本文架構…………………………………………………………………4
第二章 文獻探討……………………………………………………………5
2-1逐步設限之文獻探討……………………………………………………5
2-2 型二逐步設限之文獻探討……………………………………………5
2-3 隨機與二項移除之型二逐步設限……………………………………6
2-3-1 隨機移除之型二逐步設限…………………………………………7
2-3-2 二項移除之型二逐步設限…………………………………………7
第三章 極值分配之設限資料的統計分析…………………………………9
3-1 模式的建立………………………………………………………………9
3-1-1 二項移除之型二逐步設限…………………………………………10
3-1-2 隨機移除之型二逐步設限 …………………………………………11
3-2 參數β、θ與二項移除之機率p的估計與特性…………………………12
3-2-1參數β、θ與二項移除之機率p的最大概似估計量…………………12
3-2-2 β與θ的最大概似估計量之近似特性………………………………14
3-3 β與θ的信賴區間………………………………………………………16
3-4 期望時間………………………………………………………………21
3-5 數值計算與分析………………………………………………………25
第四章 柏拉圖分配之設限資料的統計分析 ……………………………27
4-1 模式的建立……………………………………………………………27
4-1-1 二項移除之型二逐步設限…………………………………………28
4-1-2 隨機移除之型二逐步設限…………………………………………28
4-2參數β、δ與二項移除之機率p的最大概似估計量……………………29
4-3 期望時間………………………………………………………………30
4-4 數值計算與分析………………………………………………………34
第五章 結論 ………………………………………………………………36
圖表…………………………………………………………………………38
參考文獻……………………………………………………………………77
表目錄
表3-1 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=1、θ=0.1時……………………………………………………………38
表3-2 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=-1、θ=0.1時……………………………………………………………39
表3-3 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=10、θ=0.1時…………………………………………………………40
表3-4 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=-10、θ=0.1時…………………………………………………………41
表3-5 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=1、θ=0.01時……………………………………………………………42
表3-6 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=-1、θ=0.01時…………………………………………………………43
表3-7 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=10、θ=0.01時…………………………………………………………44
表3-8 極值分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=10、θ=0.01時…………………………………………………………45
表4-1 柏拉圖分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=5、δ=1時………………………………………………………………46
表4-2 柏拉圖分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=10、δ=1時………………………………………………………………47
表4-3 柏拉圖分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=5、δ=5時………………………………………………………………48
表4-4 柏拉圖分配在二項與隨機移除之型二逐步設限下的期望時間 E[Xm]
在β=10、δ=5時………………………………………………………………49
圖目錄
圖3-1-1 極值分配β=1、θ=0.1時之期望時間的比率(REET)……………50
圖3-1-2 極值分配β=1、θ=0.1時之期望時間的比率(REET)……………51
圖3-2-1 極值分配β= -1、θ=0.1時之期望時間差的比率(|1-REET|)…52
圖3-2-2 極值分配β= -1、θ=0.1時之期望時間差的比率(|1-REET|)…53
圖3-3-1 極值分配β=10、θ=0.1時之期望時間的比率(REET)……………54
圖3-3-2 極值分配β=10、θ=0.1時之期望時間的比率(REET)……………55
圖3-4-1 極值分配β= -10、θ=0.1時之期望時間差的比率(|1-REET|)…56
圖3-4-2 極值分配β= -10、θ=0.1時之期望時間差的比率(|1-REET|)…57
圖3-5-1 極值分配β=1、θ=0.01時之期望時間的比率(REET)……………58
圖3-5-2 極值分配β=1、θ=0.01時之期望時間的比率(REET)……………59
圖3-6-1極值分配β= -1、θ=0.01時之期望時間差的比率(|1-REET|)…60
圖3-6-2極值分配β= -1、θ=0.01時之期望時間差的比率(|1-REET|)…61
圖3-7-1 極值分配β=10、θ=0.01時之期望時間的比率(REET)…………62
圖3-7-2 極值分配β=10、θ=0.01時之期望時間的比率(REET)…………63
圖3-8-1極值分配β= -10、θ=0.01時之期望時間差的比率(|1-REET|)…64
圖3-8-2極值分配β= -10、θ=0.01時之期望時間差的比率(|1-REET|)…65
圖3-9 極值分配在隨機移除之期望時間的比率(REET)……………………66
圖3-10 極值分配在隨機移除之期望時間差的比率(|1-REET|)…………67
圖4-1-1柏拉圖分配β=5、δ=1時之期望時間的比率(REET)………………68
圖4-1-2柏拉圖分配β=5、δ=1時之期望時間的比率(REET)………………69
圖4-2-1柏拉圖分配β=10、δ=1時之期望時間的比率(REET)……………70
圖4-2-2柏拉圖分配β=10、δ=1時之期望時間的比率(REET)……………71
圖4-3-1柏拉圖分配β=5、δ=5時之期望時間的比率(REET)………………72
圖4-3-2柏拉圖分配β=5、δ=5時之期望時間的比率(REET)………………73
圖4-4-1柏拉圖分配β=10、δ=5時之期望時間的比率(REET)……………74
圖4-4-2柏拉圖分配β=10、δ=5時之期望時間的比率(REET)……………75
圖4-5 柏拉圖分配在隨機移除之期望時間的比率(REET)………………76

一、中文部分：
林尚賢(民89)，在第一失敗-設限抽樣方案下Weibull分配和Extreme-value分配的參數估計，淡江大學統計學系應用統計學碩士班碩士論文。
二、英文部分：
[1] Aggarwla, R. & Balakrishan, N. (1998) Some properties of progressive censored order statistics from the pareto distribution with applications to inference, Submitted for publication.
[2] Bain, L. J. & Engelhart, M. (1992) Introduction to Probability and Mathematical Statistics, Duxbury Press.
[3] Balakrishan, N. ＆ Aggarwla, R. (2000) Progressive censoring：Theory, Methods and Applications (Boston, Birkhauser).
[4] Balakrishan, N. ＆ Sandhu, R. A. (1995) A simple simulational algorithm for generating progressive Type II cersored samples, The American Statistician, 49, pp. 229-230.
[5] Cohen, A. C. (1963) Progressively censored samples in the life testing, Technometrics, 5, pp. 327-339.
[6] Cohen, A. C. (1976) Progressively censored sampling in the three parameter log-normal distribution, Technometrics, 17, pp. 347-351.
[7] Cohen, A. C. & Norgaard, N. J. (1977) Progressively censored sampling in the three parameter gamma distribution, Technometrics, 19, pp. 333-340.
[8] Engelhart, M., Bin, L. J. & Shiue, W. K. (1986) Statistical analysis of a compound Exponential failure model. J. Statistical Computation and Simulation, 23, pp. 229-315.
[9] Gajjar, A. V. & Khatri, C. G. (1969) Progressively censored samples from log-normal and logistic distributions, Technometrics, 11, pp. 793-803.
[10] Gibbons, D. I. & Vance, L. C. (1983) Estimators for the 2-parameter Weibull distribution with progressively censored samples, IEE Transactions on Reliability, R-32, pp. 95-99.
[11] Gumbel, E. J. (1958) Statistics of Extremes, New York：Columbia University Press.
[12] Johnson, L. N. & Kotz, S. (1970) Distribution in Statistics：Continuous Univariate Distributions-Vol.1ž, Wiley, New York
[13] Laeless, J. F. (1982) Statistical Models for Lifetime Data (New York, Wiley).
[14] Mann, N. R. (1971) Best linear invariant estimation for Weibull parameters under progressive censoring, Technometrics, 13, pp. 521-533.
[15] Mcnolty, F., Doyle, J. & Hansen, E. (1980) Properties of the mixed Exponential failure process. Technometrics, 22, pp. 555-565.
[16] Nordquist, J. F. (1945) Theory of largest values, applied to earthquake magnitudes, Transactions of American Geographical Union, 26, pp. 29-31.
[17] Potter, W. D. (1949) Normalcy test of Precipitation and frequency studies of runoff on small watersheds, U. S. Department of Agriculture Technical Bulletin, No.985, Washington, DC：GPO.
[18] Rantz, S. F. & Riggs, H. C.(1949) Magnitude and frequency of floods in the Columbia River Basin, U. S. Geological Survey, Water Supply Paper1080, pp. 317-476.
[19] Thomas, D. R. ＆ Wilson, W. M. (1972) Linear order statistics estimation for the two-parameter Weibull and extreme-value distribution form Type II progressive censored samples, Technometrics, 14, pp. 679-691.
[20] Tse, S. K. ＆ Yuen, H. K.(1998)Expected experiment times for the Weibull distribution under progressive censoring with random removals, Journal of Applied statistics,25, pp. 75-83.
[21] Tse, S. K., Yang, C. ＆ Yuen, H. K. (2000) Statistical analysis of Weibull distributed lifetime data under Type II progressive censoring with binomial removals, Journal of Applied statistics, 27, pp. 1033-1043.
[22] Viveros, R. ＆ Balakrishnan, N. (1994) Interval estimation of parameters of life from progressively censored data, Technometrics, 36, pp. 84-91.
[23] Yuen, H. K. ＆ Tse, S. K. (1996) Parameters estimation for Weibull distributed lifetime under progressive censoring with random removals, Journal of Statistical Computation and Simulation, 55, pp. 57-71.