混合分配下區間設限樣本的可靠度推論研究

Reliability Inference Based on the Interval-Censored Samples of Mixture Distributions

楊瑾棋(Chin-Chi Yang)

指導教授 ： 蔡宗儒

淡江大學/商管學院/統計學系碩士班/碩士(2017年)

https://doi.org/10.6846/TKU.2017.01051

摘要

當製造商的元件來自兩個供應商，且個別供應商的元件品質可能不一致時，在混合的比例已知的條件下，本論文採用元件壽命服從韋伯分配的假設，研究在逐步型一區間設限樣本之混合韋伯分配下的參數估計問題，使用 Metropolis-Hasting 馬可夫鏈蒙地卡羅演算法得到模型的參數估計的結果，並以蒙地卡羅模擬評估估計方法的成效。通過模擬結果得到在大樣本下，本論文提出的估計結果相對穩定。

關鍵字

混合分配 ； 設限樣本 ； 貝氏估計 ； 馬可夫鏈蒙地卡羅演算法 ； 最大概似估計

並列摘要

When two suppliers supply components to a manufacturing company and the components from different suppliers could have different levels of quality, the mixed Weibull distributions is considered as the lifetime model of components. Moreover, an analytical Metropolis-Hasting Markov chain Monte Carlo procedure is proposed to estimate the model parameters. A simulation study is carried out to evaluate the performance of the proposed estimation method. Simulation results show that the proposed estimation method perform well with large samples.

並列關鍵字

Mixed distribution ； Censored sample ； Bayesian estimation ； Markov chain Monte Carlo algorithm ； Maximum likelihood estimation

參考文獻

1. Samuelson S.O and Kongerud J. (1994) Interval censoring in longitudinal data of respiratory symptoms in aluminium potroom workers: a comparison of methods. Statistics in Medicine 13: 1771-1780.

2. Sakamoto J., Teramukai S., Kiroaki N. and Yasuo O. (1997) A re-analysis of a randomized clinical trial for gastric cancer using interval censoring. Journal of Clinical Oncology 27: 445–446.

3. Sun J. (1997) Regression analysis of interval-censored failure time data. Statistics in Medicine 16: 497-504.

4. Lindsey J.C. and Ryan L. M. (1998) Tutorial in biostatistics methods for interval-censored data. Statistics in Medicine 17: 219-238.

5. Schick A. and Yu Q. (2000). Consistency of the GMLE with mixed case interval-censored data. Scandinavian Journal of Statistics 27: 45-55.