相依組件系統之貝氏可靠度分析方法主要有整合式分析、非整合式分析及簡化式非整合式分析三種。其中，整合式分析先產生系統先驗分配再透過系統故障資訊更新分配，具有低成本及實用的優點。非整合式分析使用較高成本的組件資訊來更新組件之先驗分配，再產生系統故障機率分配，花費的成本與分析複雜度都大幅提高，但卻可提供準確的結果。在實務上的另一考量則是簡化的非整合式分析，此方法忽視組件相依的情況，使用組件獨立的假設來分析系統可靠度。本研究的目的在於找出在貝氏可靠度分析中影響不同分析方法的相對誤差的關鍵因子與參數，透過Copula方法描述組件之間的相依性，並使用蒙地卡羅模擬系統失效事件的資料，最後使用分類樹來分類並預測不同情況（即不同參數組合下）最合適的分析方法以提供管理者一個可靠的決策法則。

In Bayesian reliability analysis, three approaches can be used to evaluate the reliability of a system. An aggregate analysis propagates component prior distributions first and then preforms Bayesian updates of aggregated prior distribution with system-level data, a disaggregate analysis performs Bayesian updates with component-level data first and then propagates component posterior distribution through the risk model, and the simplified disaggregate analysis also performs Bayesian updates with component-level data but assumes that the failure of the components are independent. Although the aggregate analysis and simplified disaggregate analysis are more practical and less expensive, the disaggregate analysis provides more accurate results. This work aims to identify the key factors that affect the magnitude of relative errors of three different approaches. In particular, a copula-based Bayesian reliability models for parallel system with dependent components (including failure rate and failure probability cases) were investigated by employing Monte Carlo simulation. Classification tree models were also used to provide a reliable guideline on selecting an appropriate analysis approach under different circumstances.

Ayyub, B. M. (2003). Risk analysis in engineering and economics. CRC Press.
Azaiez, M. N. and Bier, V. M. (1994). Perfect Aggregation in Bayesian Estimation for Several Realistic Reliability Models. Applied Mathematics and Computation, 64, 139 – 153.
Azaiez, M. N. and Bier, V. M. (1995). Perfect Aggregation for a Class of General Reliability Models with Bayesian Updating. Applied Mathematics and Computation, 73, 281 – 302.
Azaiez, M. N. and Bier, V. M. (1996). Aggregation Error in Bayesian Analysis of Reliability Systems. Management Sciences, 42, 516 – 528.
Bier, V. M. (1994). On the Concept of Perfect Aggregation in Bayesian Estimation. Reliability Engineering and System Safety, 46, 271 – 281.
Denton, F. T. and Mountain, D. C. (2011). Exploring the Effects of Aggregation Error in the Estimation of Consumer Demand-Elasticities. Economic Modelling, 28, 1747 – 1755.
Eryilmaz, S. and Tank, F. (2012). On Reliability Analysis of a Two-dependent-unit Series System with a Standby Unit. Applied Mathematics and Computation, 218, 7792 – 7797.
Estabrooks, A., Jo, T., and Japkowics, N. (2004). A Multiple Resampling Method for Learning from Imbalance Data Sets. Computational Intelligence, 20, 18-36.
Frey, H. C., Mokhtari, A., and Danish, T. (2003). Evaluation of Selected Sensitivity Analysis Methods based Upon Applications to Two Food Safety Process Risk Models. Dept. of Civil, Construction, and Environmental Eng., North Carolina State Univ, North Carolina.
Garrett, T. A. (2003). Aggregated versus Disaggregated Data in Regression Analysis: Implications for Inference. Economics Letters, 81, 61-65.
Guarro, S. and Yau, M. (2009). On the Nature and Practical Handling of the Bayesian Aggregation Anomaly. Reliability Engineering and System Safety, 94, 1050 – 1056.
Hamada, M. S., Wilson, A. G., Reese, C. S., and Martz, H. F. (2008). Bayesian Reliability. New York: Springer.
He, H. and Garcia, E. A. (2009). Learning from Imbalanced Data. IEEE Transactions on Knowledge and Data Engineering, 21, 1263-1284.
Johnson, V. E. (2010). Bayesian aggregation error?. International Journal of Reliability and Safety, 4(4), 359-365.
Kim, K. O. (2013). Bayesian Reliability when System and Subsystem Failure Data are Obtained in the Same Time Period. Journal of the Korean Statistical Society, 42, 95 – 103.
Limbourg, P., Kochs, H., Echtle, K., and Eusgeld, I. (2007). Reliability Prediction in Systems with Correlated Component Failures - An Approach Using Copulas. Architecture of Computing Systems (ARCS), 2007 20th International Conference.
Martz, H. F. and Waller, R. A. (1982). Bayesian Reliability Analysis. New York: Wiley.
Nelsen, R. B. (2013). An introduction to copulas (Vol. 139). Springer Science & Business Media.
Martz, H. F. (2007) Bayesian Reliability Analysis. In: Ruggeri F., Faltin F., and Kenett R. (eds) Encyclopedia of statistics in quality & reliability (pp.192-196). London: Wiley.
Mitkov, Ruslan (2003). The Oxford Handbook of Computational Linguistics. Oxford University Press
Mosleh, A., & Bier, V. (1992). On decomposition and aggregation error in estimation: some basic principles and examples. Risk Analysis, 12(2), 203-214.
Navarro, J. and Spizzichino, F. (2010). Comparisons of series and parallel systems with components sharing the same copula. Applied Stochastic Models in Business and Industry, 26, 775–791.
Navarro, J., Ruiz, J.M., Sandoval, C.J. (2005). A note on comparisons among coherent systems with dependent components using signatures. Statistics and Probability Letters, 72 179–185.
Navarro, J., Ruiz, J.M., Sandoval, C.J. (2006). Reliability Properties of System with Exchangeable Components and Exponential Conditional Distributions. Sociedad de Estad’istica e Investigaci’on Operativa Test, 15, 471 – 484.
Navarro, J., Ruiz, J.M., Sandoval, C.J. (2007). Properties of coherent systems with dependent components. Communications in Statistics – Theory and Methods, 36, 175-191.
Octavina (2015). Bayesian Reliability Analysis of Series Systems with Dependent Components (Unpublished master's thesis). National Taiwan University Science and Technology, Taipei, Taiwan.
Philipson, L. L. (1995). Anomalies in Bayesian Launch Range Safety Analysis. Reliability Engineering and System Safety, 49, 355 – 357.
Philipson, L. L. (1996). The Failure of Bayes System Reliability Inference Based on Data with Multi-Level Applicability. IEEE Transactions on Reliability, 45, 66 – 68.
Philipson, L. L. (2008). The “Bayesian Anomaly” and Its Practical Mitigation. IEEE Transactions on Reliability, 57, 171 – 173.
Qian, W., Yin, X., and Xie, L. (2010). Dependent Failure Reliability Assessment of Electronic System. Advances in Wireless Networks and Information Systems, 72, 171-176.
Quinlan, J. R. (1986). Induction of decision trees. Machine learning, 1(1), 81-106.
Quinlan, J.R. (1992). C4.5 Programs for Machine Learning. San Mateo, CA: Morgan Kaufmann.
Quinlan, J.R. (2012). Is See5/C5.0 Better Than C4.5?, [Online]. Retrieved from: http://www.rulequest.com/see5-comparison.html
Quinlan, J.R. (2015). Information on See5/C5.0 - RuleQuest Research Data Mining Tools, 2011. [Online]. Retrieved from: http://www.rulequest.com/see5-info.html
Quinlan, J.R. (2015). C5.0: An Informal Tutorial, [Online]. Retrieved from: http://www.rulequest.com/see5-unix.html
Roy, S., & Mukhopadhyay, C. (2014). Bayesian accelerated life testing under competing Weibull causes of failure. Communications in Statistics-Theory and Methods, 43(10-12), 2429-2451.
Saltelli, A., Ratto, M., Andres, T., Campolongo, F., Cariboni, J., Gatelli, D., . . . Tarantola, S. (2008). Global sensitivity analysis: the primer: John Wiley & Sons.
Tang, X. S., Li, D. Q., and Zhou, C. B. (2013). Bivariate Distribution Models Using Copulas for Reliability Analysis. Journal of Risk and Reliability.
Tomer, S. K., Singh, A. K., & Panwar, M. S. (2014). Bayesian analysis of masked series system lifetime data from a family of lifetime distributions. International Journal of System Assurance Engineering and Management, 5(4), 495-502.
Walter, M., Esch, S., & Limbourg, P. (2009, September). COBAREA: The COpula-BAsed REliability and Availability modeling environment. In Quantitative Evaluation of Systems, 2009. QEST'09. Sixth International Conference on the (pp. 119-120). IEEE.
Wong, Tzu-Tsung (2007). Perfect aggregation of Bayesian analysis on compositional data. Statistical Papers 48, 265-282.