本研究主要探討可維修系統可靠度的問題，一般而言我們稱該系統的可靠度衰退是指當系統兩相鄰錯誤發生的間隔時間持續縮短；反之，當系統兩相鄰錯誤發生的間隔時間持續延長，我們稱此系統可靠度成長。然而，一個系統執行中可能會受很多因素的影響，例如：工作環境的改善、重大策略的實行、資源的重新配置等使該系統的可靠度發生改變，其系統可靠度從成長至衰退（或衰退至成長）。上述的問題等同在不均勻布瓦松過程中探討可靠度狀態的改變，常見系統模型如：S型軟體可靠度成長模型。首先，我們從最簡單具有一個轉折點的布瓦松過程，其事件發生頻率在轉折點前後不同探討開始。對於具有一個轉折點的布瓦松過程，在其參數估計上我們使用貝氏分析方法。至於轉折點的先驗分布選擇，我們提出使用具有單峯性質的分布並以貝氏因子為比較選擇基準。研究發現具有單峯性質的轉折點先驗分布，其表現明顯優於均勻先驗分布；同時，在轉折點估計上我們提出使用有限制的最大概似估計方法，並且證明有限制的最大概似估計量具有一致性。接續，我們將系統模型推廣至具有單調性質的不均勻布瓦松過程中（例如：Power law process）。關於其錯誤再現率，我們提出使用無母數核密度估計方法。研究發現傳統無母數核密度估計方法無法保留系統可靠度的單調性質，因此我們提出使用無母數加權式核密度估計方法用於保留系統可靠度單調性質的限制。最後，我們亦以無母數加權式核密度估計方法推廣至估計S型軟體可靠度成長模型之具有單峯性質的錯誤再現率，該方法可以成功保留具有單峯限制的錯誤再現率，經模擬發現該方法有不錯的估計效果。

The main purpose of this study is to discuss the problem of reliability in repairable systems. Generally, we say that a repairable system is deteriorating (reliability decay), if the time between failures tend to get shorter with advancing age; if the time between failures tend to increase, then we say that the system is improving (reliability growth). In fact, the reliability from growth to decay (or vice versa) in a system may be affected by many factors such as improvement of the environment, strategy and resources. It is closely related to the problem of changing reliable state for a non-homogenous Poisson process. For example, the S-shaped software reliability growth model is a kind of these models. This study is focused on the detection of time at which point processes undergo abrupt changes. First, we started with the simplest model, assuming events occur according to a Poisson process, whose rate changes at an unknown change point. We used Bayesian analysis for Poisson process with a change-point. Regarding to the prior of change-point, we suggested using unimodal property. On the choices of prior, we chose Bayes factors for comparison. Our research found that it is more realistic to put a unimodal prior on it, while outlines the important feature of prior beliefs. Meanwhile, a maximum likelihood estimator for the change-point subject to a suitable constraint was proposed, which was shown to be consistent. For power law process, we used nonparametric kernel method to estimate the intensity function. Since its intensity function is monotonic, we proposed a weighted kernel estimate subject to the monotonic constraint. For an S-shaped software reliability growth model, its intensity function is unimodal. A weighted kernel estimate fitted the model appropriately, where the ordinary kernel estimate didn't guarantee the unimodality for an S-shaped growth model.

Akman, V. E. (1994). Change-point and change-curve modelling in stochastic
process and spatial statistics. Journal of Applied Statistical Science 1(4),
403–423.
Akman, V. E. and Raftery, A. E. (1986). Asymptotic inference of a change-point
poisson process. The Annals of Statistics 14, 1583–1590.
Ascher, H. and Feingold, H. (1984). Repairable System Reliability: Modeling,
Inference, Misconception and their Causes. New York: Marcel Dekker.
Bain, L. J. and Engelhardt,M. (1991). Statistical Analysis of Reliability and Lifetesting
Models 2nd. New York: Marcel Dekker.
Baker, R. D. (1996). Some new tests of the power law process. Technometrics
38(3), 256–265.
Berger, J. (1990). Robust Bayesian analysis: senesitivity to the prior. Journal Of
Statistical Planning and Inference 25, 303–328.
Berger, J. O. (1995). Statistical Decision Theory and Bayesian Analysis. 2nd.
New York: Springer-Verlag.
Berliner, L. M. and Goel, P. (1986). Incorporating partial prior information with
specified quantiles:ranges of posterior probabilities. Technical Report No.
357, Ohio State University, Dept. of Statistics.
Bowman, A. (1984). An alternative method of cross-validation for the smoothing
of density estimates. Biometrika 71, 353–360.
Carlin, B. P., Gelfand, A. E. and Smith, A. F. M. (1992). Hierararchical bayesian
analysis of change-point problems. Applied Statistics 41, 389–405.
Carlstein, E., Muller, H. G. and Siegmund, D. (1995). Change-Point problems.
IMS Lecture Notes-Monograph Series, Volume 23. Institute of Math. Statist.:
Hayward, CA.
Casella, G. and George, E. I. (1992). Explaining the gibbs sampler. The American
Statistician 46, 167–174.
Chang, Y. P. (2001). Estimation of parameters for nonhomogenous poisson
process : Software reliability with change-point model. Communication in
Statistics–Simulation 30(3), 623–635.
Chen, Y. P. and Singpurwalla, N. D. (1994). A non-gaussian kalman filter model
for tracking software reliability. Statistica Sinica 46, 535–548.
Chib, S. and Greenberg, E. (1995). Understanding the metropolis-hastings algorithm.
The American Statistician 49, 327–335.
Cline, D. and Hart, J. (1991). Kernel estimation of densities with discontinuities
or discontinuous derivatives. Statistics 22, 69–84.
Cowling, A., Hall, P. and Phillips, M. J. (1996). Bootstrap confidence regions for
the intensity of a poisson point process. Journal of the American Statistical
Association 91, 1516–1524.
Cox, D. R. and Lewis, P. A.W. (1966). The Statistical analysis of Series of events.
London: Methuen.
Crow, L. H. (1974). Reliability analysis for complex system:repairable systems.
Reliability and Biometry, 379–410.
Crowder,M. J., Kimber, A. C., Smith, R. L. and Sweeting, T. J. (1991). Statistical
Analysis of Reliability Data. London: Chapman and Hall.
DasGupta, A. and Bose, A. (1988). -minmax estimation and restricted risk
Bayes estimation of multiple Poisson means. Statistics and Decsions 6, 311–
341.
DeRobertis, L. (1978). The use of partial prior knowledge in bayesian inference.
Ph.D. dissertation, Yale University, Dept. of Statistics.
Diggle, P. and Marron, J. S. (1989). Equivalence of smoothing parameter selectors
in density and intensity estimation. Journal of the American Statistical
Association 83, 739–800.
Duane, J. T. (1964). Learning curve approach to reliability. IEEE Transaction on
Aerospace 2, 563–566.
Efron, B. and Tibshirani, R. J. (1993). An Introduction to the Bootstrap. New
York: Chapman and Hall.
Feller,W. (1971). An introduction to Probability and its application 2. New York:
2nd ed, John Wiley.
Gelfand, A. E. and Smith, A. F. M. (1990). Sampling based approaches to calculating
marginal densities. Journal of the American Statistical Association 85,
398–409.
Geman, S. and Geman, D. (1984). Stochastic relaxation, gibbs distributions, and
the bayesian restoration of images. IEEE Transactions on Pattern Analysis
and Machine Intelligence 6, 721–741.
Goel, A. L. and Okumoto, K. (1979). Time dependent error detection rate model
for software reliability and other performance measures. IEEE Transactions
on Reliability 28, 206–211.
Good, I. J. (1983). Good Thinking: The Foundations of Probability and Its Applications.
Minneapolis: University of Minnesota.
Green, P. J. (1995). Reversible jump markov chain monte carlo computation and
bayesian model determination. Biometrika 82, 711–732.
Hall, P. and Huang, L. S. (2002). Unimodal density estimation using kernel methods.
Statistica Sinica 12, 965–990.
Hall, P., Huang, L. S., Gifford, J. A. and Gijbels, I. (2001). Nonparametric estimation
of hazard rate under the constraint of monotonicity. Journal of Computational
and Graphical Statistics 10(3), 592–614.
Henderson, R. and Matthews, J. N. S. (1993). An investigation of change-points
in the annual number of cases of haemolytic uraemic syndrome. Applied
Statistics 42, 461–471.
Hinkley, D. V. (1970). Inference about the change-point in a sequence of random
variables. Biometrika 57(1), 1–17.
Hinkley, D. V. (1971). Inference about the change-point from cumulative sum
tests. Biometrika 58(3), 509–523.
Huber, P. J. (1973). Robust Statistics. New York: John Wiley.
Jarrett, R. G. (1979). A note on the intervals between coal mining disasters.
Biometrika 66, 191–193.
Khoshgoftaar, T. M. and Woodcock, T. G. (1991). Software reliability model
selection, a case study. Proc. Int. Symp. on Software Reliability Eng., 183–
191.
Krishnaiah, R. R. and Miao, B. Q. (1988). Review about estimation of changepoints
In Handbook of Statistics 7. Amsterdam: North-Holland.
Kuo, L., Lee, J. C., Choi, K. and Yang, T. Y. (1997). Bayes inference for s-shaped
software reliability growth models. IEEE Transactions on Reliability 46(1),
76–80.
Kuo, L. and Yang, T. Y. (1995). Bayesian computation of software reliability.
Journal of Computational and Graphical Statistics 4, 65–82.
Lee, C. B. (1998). Bayesian analysis of a change-point in exponential families
with application. Computational Statistics and Data Analysis 27, 195–208.
Lee, C. H., Kim, Y. T. and Park, D. H. (2004). S-shaped software reliability
growth models derived from stochastic differential equations. IIE Transactions
36, 1193–1199.
Lee, L. (1980). Testing adequacy of the weibull and log linear models for a poisson
process. Technometrics 22, 195–200.
Lin, Y. H. and Zen, M. M. (2003). Change-point estimation of a poisson process
based on arrival times. Technical Report No. 52, Department of Statistics,
National Cheng-Kung University.
Lin, Y. H. and Zen,M.M. (2007). Empirical Bayesian analysis of a change-point
Poisson process. Bulletin of Diwan College of Management 2, 1–15.
Loader, C. R. (1992). A log-linear model for a change point poisson process. The
Annals of Statistics 20, 1391–1411.
Maguire, B. A., Pearson, E. S. and Wynn, A. H. A. (1952). The time intervals
between industrial accidents. Biometrika 39, 168–180.
Meeker, W. Q. and Escobar, L. A. (1998). Statistical Methods for Reliability
Data. New York: John Wiley and Sons, Inc.
Morris, C. (1983). Parametric empirical Bayes inference: theory and applications.
Journal of the American Statistical Association 78, 47–65.
Nikora, A. P. and Lyu, M. R. (1996). Software reliability measurement experience.
New York: McGraw-Hill.
Ohba, M., Yamada, S., Takeda, K. and Osaki, S. (1982). S-shaped software reliability
growth curve:how good is it? COMPSAC’82, 38–44.
Park, W. J. and Kim, Y. G. (1992). Goodness-of-fit tests for the power law process.
IEEE Transactions on Reliability 41(1), 107–111.
Phillips, M. J. (2000). Bootstrap confidence regions for the expected rocof of a
repairable system. IEEE Transactions on Reliability 49(2), 204–208.
Phillips, M. J. (2001). Estimation of the expected rocof of a repairable system
with bootstrap confidence region. Quality and Reliability Engineering International
17, 159–162.
Raftery, A. E. and Akman, V. E. (1986). Bayesian analysis of a poisson process
with a change-point. Biometrika 73, 85–89.
Rigdon, S. E. and Basu, A. P. (1989a). Estimating the intensity function of a
weibull process at the current time: Failure truncated case. Journal of Statistical
Computation and Simulation 30, 17–38.
Rigdon, S. E. and Basu, A. P. (1989b). The power law process: A model for the
reliability of repairable system. Journal of Quality Technology 21(4), 251–
260.
Rigdon, S. E. and Basu, A. P. (2000). Statistical Methods for the Reliability of
Repairable Systems. New York: John Wiley.
Robbins, J. (1955). An empirical Bayes approach to statistcs: senesitivity to the
prior. In Proc. Third Berkeley Symp. Math. Statist. Probab. University of California
Press, Berkeley.
Rudemo, C. (1982). Empirical choice of histograms and kernel density estimators.
Scandinavian Journal of Statistics 9, 65–78.
Ruggeri, F. and Sivaganesan, S. (2005). On modelling cahnge-points in nonhomogenous
poisson processes. Statistical Inference for Stochastic Processes
8, 311–329.
Scariano, S. M. and Watkins, T. A. (1988). Nonparametric point estimators for
the change-point problem. Communication in Statistics–Theory and Methods
17(11), 3645–3675.
Schuster, E. F. (1985). Incorporating support constraints into nonparametric estimators
of densities. Communication in Statistics–Theory and Methods 14,
1123–1136.
Silverman, B. W. (1993). Density Estimation for Statistics and Data Analysis.
London: Chapman and Hall.
Sivaganesan, S. and Berger, J. (1989). Ranges of posterior measures for priors
with unimodal contaminations. The Annals of Statistics 17(2), 868–889.
Smith, A. F. M. (1975). A bayesian approach to inference about a change-point
in a sequence of random variables. Biometrika 62(2), 407–416.
Tanner, M. and Wong, W. (1987). The calculation of posterior distributions by
data augmentation. Journal of the American Statistical Association 82, 528–
550.
Tierney, L. (1994).Markov chains for exploring posterior distributions (with discussion).
The Annals of Statistics 22, 1701–1758.
West, W. R. and Ogden, R. T. (1997). Continuous-time estimation of a changepoint
in a poisson process. Journal of Statistical Computation and Simulation
56, 293–302.
Worsley, K. J. (1986). Confidence regions and test for a change-point in a sequence
of exponential family random variables. Biometrika 73, 91–104.
Yamada, S., Ohba, M. and Osaki, S. (1983). S-shaped reliability growth modeling
for software error detection. IEEE Transactions on Reliability 32, 475–
478.
Yamada, S. and Osaki, S. (1983). Reliability growth models for hardware and
software systems based on nonhomogeneous poisson process: A survey. Microelectronics
and Reliability 23, 91–112.
Yao, Y. C. (1986). Maximum likelihood estimation in hazard rate models with
a change-point. Communication in Statistics–Theory and Methods 15(8),
2455–2466.
Zen, M.M. and DasGupta, A. (1993). Estimating a Binomial parameter:is robust
Bayes real Bayes. Statistics and Decsions 11, 37–60.
Zhao, M. (1993). Change-point problems in software and hardware reliability.
Communication in Statistics–Theory and Methods 22, 757–768.