Skip to main content
Log in

A Bayesian inference to estimate change point for traffic intensity in M/M/1 queueing model

  • Application Article
  • Published:
OPSEARCH Aims and scope Submit manuscript

Abstract

The paper is concerned with the problem of change point for the inter arrival time distribution for the M/M/1 queueing system by considering the number of customers present in the system. Bayesian estimators of traffic intensities, before the change \((\rho _1)\) and after the change \((\rho _2)\), and the change point m are derived using the informative as well as non-informative priors under different loss functions. Finally a numerical example along with a practical example is given to illustrate the results.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2

Similar content being viewed by others

References

  1. Acharya, S.K., Villarreal-Rodrıguez, C.E.: Change point estimation of service rate in \(M/M/1\) queue. Int. J. Math. Oper. Res. 5(1), 110–120 (2013)

    Article  Google Scholar 

  2. Acharya, S.K., Singh, S.K., Villarreal-Rodrıguez, C.E.: Asymptotic study on change point problem for waiting time data in a single server queue. Int. J. Manag. Sci. Eng. Manag. 15(1), 39–46 (2020)

    Google Scholar 

  3. Almeida, M.A.C., Cruz, F.R.B.: A note on Bayesian estimation of traffic intensity in single-server Markovian queues. Commun. Stat. Simul. Comput. 47(9), 2577–2586 (2018)

    Article  Google Scholar 

  4. Armero, C.: Bayesian analysis of \(M/M/1/\infty /FIFO\) queues. Bayesian Statistics, 2nd edn. 613–617 (1985)

  5. Armero, C.: Bayesian inference in Markovian queues. Queueing Syst. 15, 419–426 (1994)

    Article  Google Scholar 

  6. Barry, D., Hartigan, J.A.: A Bayesian analysis for change point problems. J. Am. Stat. Assoc. 88(421), 309–319 (1993)

    Google Scholar 

  7. Basak, A., Choudhury, A.: Bayesian inference and prediction in single server M/M/1 queuing model based on queue length. Commun. Stat. Simul. Comput. (2019). https://doi.org/10.1080/03610918.2019.1586924

    Article  Google Scholar 

  8. Benes, V.E.: A sufficient set of statistics for simple telephonic exchange model. Bell Syst. Tech. J. 36, 939–964 (1957)

    Article  Google Scholar 

  9. Calabria, R., Pulcini, G.: An engineering approach to Bayes estimation for the Weibull distribution. Microelectron. Reliab. 34, 789–802 (1994)

    Article  Google Scholar 

  10. Chen, J., Gupta, A.K.: Parametric Statistical Change Point Analysis. Birkhäuser, Berlin (2000)

    Book  Google Scholar 

  11. Chernoff, H., Zacks, S.: Estimating the current mean of a normal distribution which is subjected to changes in time. Ann. Math. Stat. 35, 999–1018 (1964)

    Article  Google Scholar 

  12. Choudhury, A., Borthakur, A.C.: Bayesian inference and prediction in the single server Markovian queue. Metrika 67(3), 371–383 (2008)

    Article  Google Scholar 

  13. Choudhury, A., Basak, A.: Statistical inference on traffic intensity in an \(M/M/1\) queueing system. Int. J. Manag. Sci. Eng. Manag. 13(4), 274–279 (2018)

    Google Scholar 

  14. Chowdhury, S., Mukherjee, S.P.: Estimation of traffic intensity based on queue length in a single \(M/M/1\) queue. Commun. Stat. Theory Methods 42(13), 2376–2390 (2013)

    Article  Google Scholar 

  15. Clarke, A.B.: Maximum likelihood estimates in a simple queue. Ann. Math. Stat. 28, 1036–1040 (1957)

    Article  Google Scholar 

  16. Cruz, F.R.B., Quinino, R.C., Ho, L.L.: Bayesian estimation of traffic intensity based on queue length in a multi-server \(M/M/s\) queue. Commun. Stat. Simul. Comput. 46(9), 7319–7331 (2017)

    Article  Google Scholar 

  17. Deepthi, V., Jose, J.K.: Bayesian inference on M/M/1 queue under asymmetric loss function using Markov Chain Monte Carlo method. J. Stat. Manag. Syst. (2021). https://doi.org/10.1080/09720510.2020.1794529

    Article  Google Scholar 

  18. Gross, D., Harris, C.M.: Fundamentals of Queueing Theory, 3rd edn. Wily, New York (1998)

    Google Scholar 

  19. Hall, R.W.: Queueing Methods for Services and Manufacturing. Prentice Hall, Engelwood Cliffs (1991)

    Google Scholar 

  20. Hinkley, D.V.: Inference about the change-point in a sequence of random variables. Biometrika 57(1), 1–17 (1970)

    Article  Google Scholar 

  21. Jain, S.: Estimating changes in traffic intensity for \(M/M/1\) queueing systems. Microelectron. Reliab. 35(11), 1395–1400 (1995)

    Article  Google Scholar 

  22. Jain, S.: Estimating the change point of Erlang interarrival time distribution. INFOR 39(2), 200–207 (2001)

    Google Scholar 

  23. Kiapour, A.: Bayesian estimation of the expected queue length of a system M/M/1 with certain and uncertain priors. Commun. Stat. Theory Methods (2020). https://doi.org/10.1080/03610926.2020.1838543

    Article  Google Scholar 

  24. Lee, C.B.: Bayesian analysis of a change point in exponential families with application. Comput. Stat. Data Anal. 27, 195–208 (1998)

    Article  Google Scholar 

  25. McGrath, M.F., Gross, D., Singpurwalla, N.D.: A subjective Bayesian approach to the theory of queues \(I-\)Modelling. Queueing Syst. 1, 317–333 (1987)

    Article  Google Scholar 

  26. McGrath, M.F., Singpurwalla, N.D.: A subjective Bayesian approach to the theory of queues \(II-\)Inference and Information in \(M/M/1\) queues. Queueing Syst. 1, 335–353 (1987)

    Article  Google Scholar 

  27. Muddapur, M.V.: Bayesian estimation of parameters in some queueing models. Inst. Stat. Math. 24, 327–331 (1972)

    Article  Google Scholar 

  28. Mukherjee, S.P., Chowdhury, S.: Bayesian estimation of traffic intensity. IAPQR Trans. 30(2), 89–100 (2005)

    Google Scholar 

  29. Norstrom, J.G.: The use of precautionary loss function in risk analysis. IEEE Trans. Reliab. 45(3), 400–403 (1996)

    Article  Google Scholar 

  30. Raftery, A.E., Akman, V.E.: Bayesian analysis of a Poisson process with a change-point. Biometrika 73, 85–89 (1986)

    Article  Google Scholar 

  31. Ren, H., Wang, G.: Bayes estimation of traffic intensity in \(M/M/1\) queue under a precautionary function. Procedia Eng. 29, 3646–3650 (2012)

    Article  Google Scholar 

  32. Singh, S.K., Acharya, S.K.: Equivalence between Bayes and the maximum likelihood estimator in \(M/M/1\) queue. Commun. Stat. Theory Methods 48(19), 4780–4793 (2019a)

    Article  Google Scholar 

  33. Singh, S.K., Acharya, S.K.: Bayesian change point problem for traffic intensity in \(M/E_r/1\) queueing model. Jpn. J. Stat. Data Sci. 2, 49–70 (2019b)

    Article  Google Scholar 

  34. Singh, S.K., Acharya, S.K.: On the rate of convergence in the Bernstein-von Mises theorem for \(M/M/1\) queue. J. Indian Soc. Probab. Stat. (2021). https://doi.org/10.1007/s41096-021-00099-x

    Article  Google Scholar 

  35. Smith, A.F.M.: A Bayesian approach to inference about a change-point in a sequence of random variables. Biometrika 62, 407–416 (1975)

    Article  Google Scholar 

  36. Thiagarajan, T.R., Harris, C.M.: Statistical tests for exponential service from \(M/G/1\) waiting time data. Naval Res. Logist. Q. 26, 511–520 (1979)

    Article  Google Scholar 

  37. Thiruvaiyaru, D., Basawa, I.V.: Empirical Bayes estimation for queueing systems and networks. Queueing Syst. 11, 179–202 (1992)

    Article  Google Scholar 

  38. Vaidyanathan, V.S., Chandrasekhar, P.: Parametric estimation of an \(M/E_r/1\) queue. OPSEARCH (2018). https://doi.org/10.1007/s12597-018-0342-0

    Article  Google Scholar 

  39. Wolfram Research Inc.: Generalized incomplete beta function (2015). http://functions.wolfram.com/GammaBetaErf/Beta4/. Accessed on 09 Nov 2015

Download references

Acknowledgements

We are thankful to the referees for their comments and useful suggestions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Saroja Kumar Singh.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

1.1 Derivation of joint and marginal posterior distributions of m, \(\rho _1\) and \(\rho _2\) under different prior

Beta Prior

$$\begin{aligned}&\pi _1(\rho _1,\rho _2,m|x) = \frac{\prod _{i=1}^{m}p(x_i)\pi _1(\rho _1) \cdot \prod _{i=m+1}^{n}p(x_i)\pi _1(\rho _2)}{ \sum _{m=1}^{n-1}\int _{0}^{1} \int _{0}^{1} \left[ \prod _{i=1}^{m}p(x_i)\pi _1(\rho _1) \cdot \prod _{i=m+1}^{n}p(x_i)\pi _1(\rho _2) \right] d{\rho _1} d{\rho _2} } \\&\quad = \frac{ \frac{ \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1}}{{\mathscr {B}}(a_1, b_1 )} \cdot \frac{ \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} }{{\mathscr {B}}(a_2, b_2 )} }{ \sum _{m=1}^{n-1} \int _{\rho _1} \int _{\rho _2} \frac{ \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1}}{{\mathscr {B}}(a_1, b_1 )} \cdot \frac{ \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} }{{\mathscr {B}}(a_2, b_2 )} d\rho _1 d\rho _2 } \\&\quad = \frac{ \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} \cdot \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} }{ \sum _{m=1}^{n-1} \int _{0}^{1} \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} d\rho _2 } \\&\quad = \frac{ \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} \cdot \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 ) } \end{aligned}$$
$$\begin{aligned} \pi _1(m|x)&= \int _{0}^{1} \int _{0}^{1} \pi _1(\rho _1,\rho _2,m|x) d\rho _1 d\rho _2 \\&= \frac{\int _{0}^{1} \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1}d\rho _2}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\&= \frac{ {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\ \pi _1(\rho _1|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _1(\rho _1,\rho _2,m|x)d\rho _2 \\&=\frac{\sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1} d\rho _2 \right) \cdot \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\ \pi _1(\rho _2|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _1(\rho _1,\rho _2,m|x)d\rho _1 \\&=\frac{\sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _1^{S_{m}+a_1-1} (1-\rho _1)^{m+b_1-1} d\rho _1\right) \cdot \rho _2^{S_{n-m}+a_2-1} (1-\rho _2)^{n-m+b_2-1}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\&=\frac{\sum _{m=1}^{n-1}{\mathscr {B}}( S_{m}+a_1, m+b_1 ) \rho _1^{S_{n-m}+a_2-1} (1-\rho _1)^{n-m+b_2-1} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \end{aligned}$$

Truncated Uniform Prior

$$\begin{aligned} \pi _2(\rho _1,\rho _2,m|x)&= \frac{\prod _{i=1}^{m}p(x_i)\pi _2(\rho _1) \cdot \prod _{i=m+1}^{n}p(x_i)\pi _2(\rho _2)}{\sum _{m=1}^{n-1}\int _{c_1}^{d_1} \int _{c_2}^{d_2} \left[ \prod _{i=1}^{m}p(x_i)\pi _2(\rho _1)\cdot \prod _{i=m+1}^{n}p(x_i)\pi _2(\rho _2) \right] d{\rho _1} d{\rho _2}} \\&= \frac{ \rho _1^{S_{m}} (1-\rho _1)^{m} \cdot \rho _2^{S_{n-m}} (1-\rho _2)^{n-m} }{ \sum _{m=1}^{n-1} \int _{0}^{1} \rho _1^{S_{m}} (1-\rho _1)^{m} d\rho _1 \cdot \int _{0}^{1} \rho _2^{S_{n-m}} (1-\rho _2)^{n-m} d\rho _2 } \\&= \frac{\rho _1^{S_{m}} (1-\rho _1)^{m} ~\rho _2^{S_{n-m}} (1-\rho _2)^{n-m}}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \\ \pi _2(m|x)&= \int _{c_1}^{d_1} \int _{c_2}^{d_2} \pi _2(\rho _1,\rho _2,m|x) d\rho _1 d\rho _2 \\&= \frac{ \int _{c_1}^{d_1} \rho _1^{S_{m}} (1-\rho _1)^{m}d\rho _1 \cdot \int _{c_2}^{d_2}\rho _2^{S_{n-m}} (1-\rho _2)^{n-m} d\rho _2 }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \\&= \frac{ {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\ \pi _2(\rho _1|x)&= \sum _{m=1}^{n-1} \int _{c_2}^{d_2} \pi _2(\rho _1,\rho _2,m|x)d\rho _2 \\&= \frac{\sum _{m=1}^{n-1} \left( \int _{c_2}^{d_2} \rho _2^{S_{n-m}} (1-\rho _2)^{n-m} d\rho _2\right) \cdot \rho _1^{S_{m}} (1-\rho _1)^{m} }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\&= \frac{ { {\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) }} \rho _1^{S_{m}} (1-\rho _1)^{m} }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\ \pi _2(\rho _2|x)&= \sum _{m=1}^{n-1} \int _{c_1}^{d_1} \pi _2(\rho _1,\rho _2,m|x)d\rho _1 \\&= \frac{\sum _{m=1}^{n-1}\left( \int _{c_1}^{d_1} \rho _1^{S_{m}} (1-\rho _1)^{m}d\rho _1 \right) \cdot \rho _2^{S_{n-m}} (1-\rho _2)^{n-m}}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) } \rho _2^{S_{n-m}} (1-\rho _2)^{n-m} }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \end{aligned}$$

Quasi Prior

$$\begin{aligned} \pi _3(\rho _1,\rho _2,m|x)&= \frac{\prod _{i=1}^{m}p(x_i)\pi _3(\rho _1)\cdot \prod _{i=m+1}^{n}p(x_i)\pi _3(\rho _2)}{\sum _{m=1}^{n-1}\int _{\rho _1} \int _{\rho _2} \left[ \prod _{i=1}^{m}p(x_i)\pi _3(\rho _1) \cdot \prod _{i=m+1}^{n}p(x_i)\pi _3(\rho _2) \right] d{\rho _1} d{\rho _2}} \\&= \frac{ \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m} \rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m} }{ \sum _{m=1}^{n-1} \int _{01}^{1}\rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m} d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m}d\rho _2 } \\&= \frac{ \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m} \rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\ \pi _3(m|x)&= \int _{0}^{1} \int _{0}^{1} \pi _3(\rho _1,\rho _2,m|x) d\rho _1 d\rho _2 \\&=\frac{\int _{0}^{1} \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m} d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m} d\rho _2}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \\&= \frac{ {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \\ \pi _3(\rho _1|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _3(\rho _1,\rho _2,m|x)d\rho _2 \\&= \frac{\sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m} d\rho _2\right) \cdot \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\ \pi _3(\rho _2|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _3(\rho _1,\rho _2,m|x)d\rho _1 \\&=\frac{ \sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _1^{S_{m}-\alpha _1} (1-\rho _1)^{m}d\rho _1\right) \cdot \rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\&=\frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 )\rho _2^{S_{n-m}-\alpha _2} (1-\rho _2)^{n-m} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \end{aligned}$$

Jeffreys Prior

$$\begin{aligned} \pi _4(\rho _1,\rho _2,m|x)&= \frac{\prod _{i=1}^{m}p(x_i)\pi _4(\rho _1)\cdot \prod _{i=m+1}^{n}p(x_i)\pi _4(\rho _2)}{\sum _{m=1}^{n-1}\int _{\rho _1} \int _{\rho _2} \left[ \prod _{i=1}^{m}p(x_i)\pi _4(\rho _1) \cdot \prod _{i=m+1}^{n}p(x_i)\pi _4(\rho _2)\right] d{\rho _1} d{\rho _2}} \\&=\frac{\rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1} \rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1} }{ \sum _{m=1}^{n-1} \int _{0}^{1}\rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1}d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1}d\rho _2 } \\&= \frac{ \rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1} \rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\ \pi _4(m|x)&= \int _{0}^{1} \int _{0}^{1} \pi _4(\rho _1,\rho _2,m|x) d\rho _1 d\rho _2 \\&= \frac{\int _{0}^{1} \rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1} d\rho _1 \cdot \int _{0}^{1}\rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1} d\rho _2 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\&= \frac{ {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\ \pi _4(\rho _1|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _4(\rho _1,\rho _2,m|x)d\rho _2 \\&= \frac{\sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1} d\rho _2 \right) \cdot \rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+1/2, n-m )} \rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\ \pi _4(\rho _2|x)&= \sum _{m=1}^{n-1} \int _{0}^{1} \pi _4(\rho _1,\rho _2,m|x)d\rho _1 \\&= \frac{\sum _{m=1}^{n-1} \left( \int _{0}^{1} \rho _1^{S_{m}+1/2-1} (1-\rho _1)^{m -1} d\rho _1 \right) \cdot \rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m -1}}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\&=\frac{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) \rho _2^{S_{n-m}+1/2-1} (1-\rho _2)^{n-m-1} }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \end{aligned}$$

1.2 Derivation of Bayesian estimators of m, \(\rho _1\) and \(\rho _2\) under different loss functions

1.3 Squared error loss function (SELF)

Beta Prior

$$\begin{aligned} {\widehat{m}}^B_{BS}&= E(m|x) = \sum _{m=1}^{n-1} m~\pi _1(m|x)\\&= \frac{ \sum _{m=1}^{n-1} m ~{\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\ {\widehat{\rho }}^B_{1BS}&= E(\rho _1|x) = \int _{0}^{1} \rho _1~\pi _1(\rho _1|x) d\rho _1 \\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}(S_{n-m}+a_2, n-m+b_2)} \cdot \int _{0}^{1} \rho _1^{S_{m}+a_1} (1-\rho _1)^{m+b_1-1} d\rho _1 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1+1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\ {\widehat{\rho }}^B_{2BS}&= E(\rho _2|x) = \int _{0}^{1} \rho _2~\pi _1(\rho _2|x) d\rho _2\\&= \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+a_1, m+b_1)} \cdot \int _{0}^{1} \rho _2^{S_{n-m}+a_2} (1-\rho _2)^{n-m+b_2-1} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2+1, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \end{aligned}$$

Truncated Uniform Prior

$$\begin{aligned} {\widehat{m}}^T_{BS}&= E(m|x) = \sum _{m=1}^{n-1} m~\pi _2(m|x) \\&= \frac{ \sum _{m=1}^{n-1} m ~{\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \\ {\widehat{\rho }}^T_{1BS}&= E(\rho _1|x) = \int _{c_1}^{d_1} \rho _1~\pi _2(\rho _1|x) d\rho _1 \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_2, d_2)}(S_{n-m} +1, n-m+1) \cdot \int _{c_1}^{d_1} \rho _1^{S_{m} +1} (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} (S_{m}+1, m+1) {\mathscr {B}}_(c_2, d_2) ( S_{n-m} +1, n-m+1 ) } \\&= \frac{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +2, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\ {\widehat{\rho }}^T_{2BS}&= E(\rho _2|x) = \int _{c_2}^{d_2} \rho _2~\pi _2(\rho _2|x) d\rho _2 \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} (S_{m}+1, m+1) \cdot \int _{c_2}^{d_2} \rho _2^{S_{n-m}+1} (1-\rho _2)^{n-m} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \\&= \frac{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +2, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \end{aligned}$$

Quasi Prior

$$\begin{aligned} {\widehat{m}}^Q_{BS}&= E(m|x) = \sum _{m=1}^{n-1} m~\pi _3(m|x) \\&= \frac{ \sum _{m=1}^{n-1} m ~{\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \\ {\widehat{\rho }}^Q_{1BS}&= E(\rho _1|x) = \int _{0}^{1} \rho _1~\pi _3(\rho _1|x) d\rho _1 \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{n-m}-\alpha _2+1, n-m+1) \cdot \int _{0}^{1} \rho _1^{S_{m}-\alpha _1 +1} (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\&= \frac{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+2, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \\ {\widehat{\rho }}^Q_{2BS}&= E(\rho _2|x) = \int _{0}^{1} \rho _2~\pi _3(\rho _2|x) d\rho _2 \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}-\alpha _1+1, m+1) \cdot \int _{0}^{1} \rho _2^{S_{n-m}-\alpha _2+1} (1-\rho _2)^{n-m} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \\&= \frac{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+2, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \end{aligned}$$

Jeffreys Prior

$$\begin{aligned} {\widehat{m}}^J_{BS}&= E(m|x) = \sum _{m=1}^{n-1} m~\pi _1(m|x) \\&= \frac{ \sum _{m=1}^{n-1} m ~{\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\ {\widehat{\rho }}^J_{1BS}&= E(\rho _1|x) = \int _{0}^{1} \rho _1~\pi _1(\rho _1|x) d\rho _1 \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{n-m}+1/2, n-m) \cdot \int _{0}^{1} \rho _1^{S_{m}+3/2-1} (1-\rho _1)^{m -1} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+3/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\ {\widehat{\rho }}^J_{2BS}&= E(\rho _2|x) = \int _{0}^{1} \rho _2~\pi _1(\rho _2|x) d\rho _2 \\&= \frac{\sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+1/2, m) \cdot \int _{0}^{1} \rho _2^{S_{n-m}+3/2-1} (1-\rho _2)^{n-m-1} d\rho _2}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \\&= \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+3/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \end{aligned}$$

1.4 Precautionary loss function (PLF)

Beta Prior

$$\begin{aligned} {\widehat{m}}^B_{BP}&= \left[ E(m^2|x)\right] ^{\frac{1}{2}} = \left[ \sum _{m=1}^{n-1} m^2 ~\pi _1(m|x) \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^2 ~{\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^B_{1BP}&= \left[ E(\rho ^2_1|x)\right] ^{\frac{1}{2}} = \left[ \int _{0}^{1} \rho _1^2 ~\pi _1(\rho _1|x) d\rho _1 \right] ^{\frac{1}{2}} \\&=\left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 ) \cdot \int _{0}^{1} \rho _1^{S_{m}+a_1+1} (1-\rho _1)^{m+b_1-1} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1+2, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^B_{2BP}&=\left[ E(\rho _2^2|x)\right] ^{\frac{1}{2}} = \left[ \int _{0}^{1} \rho _2^2 ~\pi _1(\rho _2|x) d\rho _2 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+a_1, m+b_1)\cdot \int _{0}^{1} \rho _2^{S_{n-m}+a_2+1} (1-\rho _2)^{n-m+b_2-1} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{\frac{1}{2} } \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2+2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{\frac{1}{2}} \end{aligned}$$

Truncated Uniform Prior

$$\begin{aligned} {\widehat{m}}^T_{BP}&= \left[ E(m^2|x)\right] ^{\frac{1}{2}} = \left[ \sum _{m=1}^{n-1} m^2 ~\pi _2(m|x) \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^2 ~{\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^T_{1BP}&= \left[ E(\rho ^2_1|x)\right] ^{\frac{1}{2}} = \left[ \int _{c_1}^{d_1} \rho _1^2 ~\pi _2(\rho _1|x) d\rho _1 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )\cdot \int _{c_1}^{d_1} \rho _1^{S_{m} +2} (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \right] ^{\frac{1}{2}} \\&= \left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +3, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^T_{2BP}&= \left[ E(\rho ^2_2|x)\right] ^{\frac{1}{2}} =\left[ \int _{c_2}^{d_2} \rho _2^2 ~\pi _2(\rho _2|x) d\rho _2 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} (S_{m}+1, m+1) \cdot \int _{c_2}^{d_2} \rho _2^{S_{n-m}+2} (1-\rho _2)^{n-m} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} (S_{m}+1, m+1) {\mathscr {B}}_{(c_2, d_2)} (S_{n-m}+1, n-m+1)} \right] ^{\frac{1}{2}} \\&=\left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +3, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \right] ^{\frac{1}{2}} \end{aligned}$$

Quasi Prior

$$\begin{aligned} {\widehat{m}}^Q_{BP}&= \left[ E(m^2|x)\right] ^{\frac{1}{2}} =\left[ \sum _{m=1}^{n-1} m^2 ~\pi _3(m|x) \right] ^{\frac{1}{2}} \\&=\left[ \frac{ \sum _{m=1}^{n-1} m^2 ~ {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^Q_{1BP}&= \left[ E(\rho ^2_1|x)\right] ^{\frac{1}{2}} = \left[ \int _{0}^{1} \rho _1^2 ~\pi _3(\rho _1|x) d\rho _1 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \cdot \int _{0}^{1} \rho _1^{S_{m}-\alpha _1 +2} (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+3, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^Q_{2BP}&=\left[ E(\rho _2^2|x)\right] ^{\frac{1}{2}} = \left[ \int _{0}^{1} \rho _2^2 ~\pi _3(\rho _2|x) d\rho _2 \right] ^{\frac{1}{2}} \\&=\left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}-\alpha _1+1, m+1)\cdot \int _{0}^{1} \rho _2^{S_{n-m}-\alpha _2+2} (1-\rho _2)^{n-m} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}-\alpha _1+1, m+1) {\mathscr {B}}(S_{n-m}-\alpha _2+1, n-m+1) }\right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+3, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{\frac{1}{2}} \end{aligned}$$

Jeffreys Prior

$$\begin{aligned} {\widehat{m}}^J_{BP}&= \left[ E(m^2|x)\right] ^{\frac{1}{2}} = \left[ \sum _{m=1}^{n-1} m^2 ~ \pi _4(m|x) \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^2 ~{\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^J_{1BP}&= \left[ E(\rho ^2_1|x)\right] ^{\frac{1}{2}}= \left[ \int _{0}^{1} \rho _1^2 ~\pi _4(\rho _1|x) d\rho _1 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+1/2, n-m )} \cdot \int _{0}^{1} \rho _1^{S_{m}+5/2-1} (1-\rho _1)^{m -1} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+5/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{\frac{1}{2}} \\ {\widehat{\rho }}^J_{2BP}&=\left[ E(\rho ^2_1|x)\right] ^{\frac{1}{2}} =\left[ \int _{0}^{1} \rho _1^2 ~\pi _4(\rho _1|x) d\rho _1 \right] ^{\frac{1}{2}} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+1/2, m) \cdot \int _{0}^{1} \rho _2^{S_{n-m}+5/2-1} (1-\rho _2)^{n-m-1} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+1/2, m) {\mathscr {B}}(S_{n-m}+1/2, n-m )} \right] ^{\frac{1}{2}}\\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+5/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{\frac{1}{2}} \end{aligned}$$

1.5 General entropy loss function (GELF)

Beta Prior

$$\begin{aligned} {\widehat{m}}^B_{BE}&= \left[ E(m^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \sum _{m=1}^{n-1} m^{-\gamma } ~\pi _1(m|x) \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^{-\gamma } ~{\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^B_{1BE}&= \left[ E(\rho _1^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{0}^{1} \rho _1^{-\gamma } ~\pi _1(\rho _1|x) d\rho _1 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \cdot \int _{0}^{1} \rho _1^{S_{m}+a_1-\gamma -1} (1-\rho _1)^{m+b_1-1} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1-\gamma , m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^B_{2BE}&=\left[ E(\rho _2^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{0}^{1} \rho _2^{-\gamma }~\pi _1(\rho _2|x) d\rho _2 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}+a_1, m+b_1) \cdot \int _{0}^{1} \rho _2^{S_{n-m}+a_2-\gamma -1} (1-\rho _2)^{n-m+b_2-1} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2-\gamma , n-m+b_2 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+a_1, m+b_1 ) {\mathscr {B}}( S_{n-m}+a_2, n-m+b_2 )} \right] ^{-\frac{1}{\gamma }} \end{aligned}$$

Truncated Uniform Prior

$$\begin{aligned} {\widehat{m}}^T_{BE}&= \left[ E(m^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \sum _{m=1}^{n-1} m^{-\gamma } ~\pi _2(m|x) \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^{-\gamma } ~{\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^T_{1BE}&= \left[ E(\rho _1^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} =\left[ \int _{c_1}^{d_1} \rho _1^{-\gamma } ~\pi _2(\rho _1|x) d\rho _1 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \cdot \int _{c_1}^{d_1} \rho _1^{S_{m} - \gamma } (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) } \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} -\gamma +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^T_{2BE}&= \left[ E(\rho _2^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{c_2}^{d_2} \rho _2^{-\gamma } ~\pi _2(\rho _2|x) d\rho _2 \right] ^{-\frac{1}{\gamma }} \\&=\left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m}+1, m+1) \cdot \int _{c_2}^{d_2} \rho _2^{S_{n-m}-\gamma } (1-\rho _2)^{n-m} d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1)} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m}+1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} -\gamma +1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}_{(c_1, d_1)} ( S_{m} +1, m+1 ) {\mathscr {B}}_{(c_2, d_2)} ( S_{n-m} +1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \end{aligned}$$

Quasi Prior

$$\begin{aligned} {\widehat{m}}^Q_{BE}&= \left[ E(m^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \sum _{m=1}^{n-1} m^{-\gamma } ~\pi _3(m|x) \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^{-\gamma } ~{\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^Q_{1BE}&=\left[ E(\rho _1^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{0}^{1} \rho ^{-\gamma } ~\pi _3(\rho _1|x) d\rho _1 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \cdot \int _{0}^{1} \rho _1^{S_{m}-\alpha _1 - \gamma } (1-\rho _1)^{m} d\rho _1 }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1-\gamma +1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 ) } \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^Q_{2BE}&= \left[ E(\rho _2^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{0}^{1} \rho _2^{-\gamma } ~\pi _3(\rho _2|x) d\rho _2 \right] ^{-\frac{1}{\gamma }} \\&=\left[ \frac{\sum _{m=1}^{n-1} {\mathscr {B}}(S_{m}-\alpha _1+1, m+1) \cdot \int _{0}^{1} \rho _2^{S_{n-m}-\alpha _2-\gamma } (1-\rho _2)^{m}d\rho _2 }{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2-\gamma +1, n-m+1 )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\alpha _1+1, m+1 ) {\mathscr {B}}( S_{n-m}-\alpha _2+1, n-m+1 )} \right] ^{-\frac{1}{\gamma }} \end{aligned}$$

Jeffreys Prior

$$\begin{aligned} {\widehat{m}}^J_{BE}&= \left[ E(m^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \sum _{m=1}^{n-1} m^{-\gamma } ~\pi _4(m|x) \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} m^{-\gamma } ~{\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^J_{1BE}&= \left[ E(\rho _1^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} = \left[ \int _{0}^{1} \rho _1^{-\gamma } ~\pi _4(\rho _1|x) d\rho _1 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ {\sum _{m=1}^{n-1} {\mathscr {B}}( S_{n-m}+1/2, n-m )} \cdot \int _{0}^{1} \rho _1^{S_{m}-\gamma +1/2-1} (1-\rho _1)^{m -1} d \rho _1}{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{- \frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}-\gamma +1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{-\frac{1}{\gamma }} \\ {\widehat{\rho }}^J_{2BE}&= \left[ E(\rho _2^{-\gamma }|x)\right] ^{-\frac{1}{\gamma }} =\left[ \int _{0}^{1} \rho _2^{-\gamma } ~\pi _4(\rho _2|x) d\rho _2 \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{\sum _{m=1}^{n-1}{\mathscr {B}}(S_{m}+1/2, m) \cdot \int _{0}^{1} \rho _2^{S_{n-m}-\gamma +1/2-1} (1-\rho _2)^{n-m-1} d\rho _2 }{ \sum _{m=1}^{n-1}{\mathscr {B}}(S_{m}+1/2, m) {\mathscr {B}}(S_{n-m}+1/2, n-m )} \right] ^{-\frac{1}{\gamma }} \\&= \left[ \frac{ \sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}-\gamma +1/2, n-m ) }{\sum _{m=1}^{n-1} {\mathscr {B}}( S_{m}+1/2, m ) {\mathscr {B}}( S_{n-m}+1/2, n-m )} \right] ^{-\frac{1}{\gamma }} \end{aligned}$$

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Singh, S.K., Acharya, S.K. A Bayesian inference to estimate change point for traffic intensity in M/M/1 queueing model. OPSEARCH 59, 166–206 (2022). https://doi.org/10.1007/s12597-021-00535-3

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12597-021-00535-3

Keywords

Mathematics Subject Classification

Navigation