Skip to main content

Exact Likelihood-Based Point and Interval Estimation for Lifetime Characteristics of Laplace Distribution Based on hybrid Type-I and Type-II Censored Data

  • Conference paper
  • First Online:
Trends and Perspectives in Linear Statistical Inference

Part of the book series: Contributions to Statistics ((CONTRIB.STAT.))

Abstract

In this chapter, we first derive explicit expressions for the Maximum likelihood estimators (MLEs) of the parameters of Laplace distribution-based on a hybrid Type-I censored sample (Type-I HCS). We then derive the conditional moment generating functions (MGF) of the MLEs, and then use them to obtain the means, variances, and covariance of the MLEs. From the conditional MGFs, we also derive the exact conditional distributions of the MLEs, which are then used to develop exact conditional confidence intervals (CIs) for the parameters. Proceeding similarly, we obtain the MLEs of quantile, reliability, and cumulative hazard functions, and discuss the construction of exact CIs for these functions as well. By using the relationships between Type-I, Type-II, Type-I HCS, and hybrid Type-II censored samples (Type-II HCS), we develop exact inferential methods based on a Type-II HCS as well. Then, a Monte Carlo simulation study is carried out to evaluate the performance of the developed inferential results. Finally, a numerical example is presented to illustrate the point and interval estimation methods developed here under both Type-I HCS and Type-II HCS.

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

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 84.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 109.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 109.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

References

  • Arnold, B.C., Balakrishnan, N., Nagaraja, N.H.: A First Course in Order Statistics. Wiley, New York (1992)

    MATH  Google Scholar 

  • Bain, L.J., Engelhardt, M.: Interval estimation for the two-parameter double exponential distribution. Technometrics 15, 875–887 (1973)

    Article  MathSciNet  MATH  Google Scholar 

  • Balakrishnan, N., Chandramouleeswaran, M.P.: Reliability estimation and tolerance limits for Laplace distribution based on censored samples. Mircoelectronics Reliab. 36, 375–378 (1996)

    Article  Google Scholar 

  • Balakrishnan, N., Cohen, A.C.: Order Statistics and Inference: Estimation Methods. Academic Press, Boston (1991)

    MATH  Google Scholar 

  • Balakrishnan, N., Cutler, C.D.: Maximum likelihood estimation of Laplace parameters based on Type-II censored samples. In: Nagaraja, H.N., Sen, P.K., Morrison, D.F. (eds.) Statistical Theory and Applications: Papers in Honor of Herbert A. David, pp. 145–151. Springer, New York (1995)

    Google Scholar 

  • Balakrishnan, N., Kundu, D.: Hybrid censoring: models, inferential results and applications. Comput. Stat. Data Anal. 57, 166–209 (2013) (with discussions)

    Google Scholar 

  • Balakrishnan, N., Zhu, X.: Exact likelihood-based point and interval estimation for Laplace distribution based on Type-II right censored samples. J. Stat. Comput. Simul. 86, 29–54 (2016)

    Article  MathSciNet  Google Scholar 

  • Childs, A., Balakrishnan, N.: Conditional inference procedures for the Laplace distribution based on Type-II right censored samples. Stat. Probab. Lett. 31, 31–39 (1996)

    Article  MathSciNet  MATH  Google Scholar 

  • Childs, A., Balakrishnan, N.: Maximum likelihood estimation of Laplace parameters based on general Type-II censored samples. Stat. Pap. 38, 343–349 (1997)

    Article  MATH  Google Scholar 

  • Childs, A., Balakrishnan, N.: Conditional inference procedures for the Laplace distribution when the observed samples are progressively censored. Metrika 52, 253–265 (2000)

    Article  MathSciNet  MATH  Google Scholar 

  • Childs, A., Chandrasekar, B., Balakrishnan, N., Kundu, D.: Exact likelihood inference based on Type-I and Type-II hybrid censored samples from the exponential distribution. Ann. Inst. Stat. Math. 55, 319–330 (2003)

    MathSciNet  MATH  Google Scholar 

  • Devore, J.: Probability and Statistics for Engineering and the Sciences. Cengage Learning (2015)

    Google Scholar 

  • Epstein, B.: Truncated life tests in the exponential case. Ann. Math. Stat. 25, 555–564 (1954)

    Article  MathSciNet  MATH  Google Scholar 

  • Iliopoulos, G., Balakrishnan, N.: Exact likelihood inference for Laplace distribution based on Type-II censored samples. J. Stat. Plan. Inference 141, 1224–1239 (2011)

    Article  MathSciNet  MATH  Google Scholar 

  • Iliopoulos, G., MirMostafaee, S.M.T.K.: Exact prediction intervals for order statistics from the Laplace distribution based on the maximum-likelihood estimators. Statistics 48, 575–592 (2014)

    Article  MathSciNet  MATH  Google Scholar 

  • Johnson, N.L., Kotz, S., Balakrishnan, N.: Continuous Univariate Distributions, vol. 1, 2nd edn. Wiley, New York (1994)

    Google Scholar 

  • Johnson, N.L., Kotz, S., Balakrishnan, N.: Continuous Univariate Distributions, vol. 2, 2nd edn. Wiley, New York (1995)

    Google Scholar 

  • Kappenman, R.F.: Conditional confidence intervals for double exponential distribution parameters. Technometrics 17, 233–235 (1975)

    Article  MathSciNet  MATH  Google Scholar 

  • Kappenman, R.F.: Tolerance intervals for the double exponential distribution. J. Am. Stat. Assoc. 72, 908–909 (1977)

    MathSciNet  MATH  Google Scholar 

  • Kotz, S., Kozubowski, T.J., Podgorski, K.: The Laplace Distribution and Generalizations: A Revisit with Applications to Communications, Economics, Engineering, and Finance. Wiley, New York (2001)

    Book  MATH  Google Scholar 

  • Puig, P., Stephens, M.A.: Tests of fit for the Laplace distribution, with applications. Technometrics 42, 417–424 (2000)

    Article  MathSciNet  MATH  Google Scholar 

  • Zhu, X., Balakrishnan, N.: Exact inference for Laplace quantile, reliability and cumulative hazard functions based on Type-II censored data. IEEE Trans. Reliab. 65, 164–178 (2016)

    Article  Google Scholar 

  • Zhu, X., Balakrishnan, N.: Exact likelihood-based point and interval estimation for lifetime characteristics of laplace distribution based on a time-constrained life-testing experiment. In: Adhikari, A., Adhikari, M.R., Chaubey, Y.P. (eds.) Mathematical and Statistical Applications in Life Sciences and Engineering, Accepted. Springer, New York (2017) (to appear)

    Google Scholar 

Download references

Acknowledgements

This research was partially supported by the National Natural Science Foundation of China (No. 11571263) and by the National Sciences and Engineering Research Council of Canada.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to N. Balakrishnan .

Editor information

Editors and Affiliations

Appendix

Appendix

List of notation used throughout

$$\begin{aligned} p_1= & {} \frac{(-1)^ln!e^{-\frac{(T-\mu )(n+l-d)}{\sigma }}}{2^n(n-d)!j!(d-j-l)!l!(1-p_0)},\\ p_{2,a}= & {} \frac{(-1)^ln!}{2^n(n-k)!j!(k-j-1-l)!l!(n-k)!(n-k+l+1)(1-p_0)},\\ p_{2,b}= & {} -p_{2,a}e^{-\frac{(T-\mu )(n-k+l+1)}{\sigma }},\\ p_{3}= & {} \frac{(-1)^l n!}{2^{k+l}(k-1)!(n-k-l)!l!(k+l)(1-p_0)},\\ p_{4,a,e}= & {} \frac{(-1)^l n!e^{-\frac{(T-\mu )m}{\sigma }}}{2^n j!(m-1-j-l)!(l+1)!m!(1-p_0)},\\ p_{4,b,e}= & {} -p_{4,a,e}e^{-\frac{(T-\mu )(l+1)}{\sigma }},\\ p_{5,e}= & {} \frac{n!e^{-\frac{(T-\mu )m}{\sigma }}}{2^n (n-m)! m!(1-p_0)},\\ p_{6,a,e}= & {} \frac{(-1)^ln!}{2^nj!(m-1-j-l)!m!(l+1)!(1-p_0)},\\ p_{6,b,e}= & {} -p_{6,a,e}e^{-\frac{(T-\mu )m}{\sigma }},\\ p_{6,c,e}= & {} \frac{(-1)^{l+1}n!}{2^nj!(m-1-j-l)!(m-1)!(l+1)!(m+l+1)(1-p_0)},\\ p_{6,d,e}= & {} -p_{6,c,e}e^{-\frac{(T-\mu )(m+l+1)}{\sigma }},\\ p_{7,a,e}= & {} \frac{n!}{2^nm!m!(1-p_0)},\\ p_{7,b,e}= & {} -p_{7,a,e}e^{-\frac{m(T-\mu )}{\sigma }},\\ p_{9,e}= & {} \frac{(-1)^l n!}{m!(n-k-l)!l!2^{k+l}(m+k+l)(1-p_0)},\\ p_{10,a,e}= & {} \frac{(-1)^{l_1+d-m-1-l_2}n!e^{-\frac{(T-\mu )(m-1-l_2)}{\sigma }}}{2^nj!(m-1-j-l_1)!(d-m-1-l_2)!(l_2+1)!(n-d)!(l_1+1)!(1-p_0)},\\ p_{10,b,e}= & {} -p_{10,a,e}e^{-\frac{(T-\mu )(l_2+1)}{\sigma }},\\ p_{10,c,e}= & {} -\frac{p_{10,a,e}}{l_1+l_2+2},\\ p_{10,d,e}= & {} -p_{10,c,e}e^{-\frac{(T-\mu )(l_1+l_2+2)}{\sigma }},\\ p_{11,a,e}= & {} \frac{(-1)^{d-m-l-1}n!e^{-\frac{(T-\mu )(m-l-1)}{\sigma }}}{2^nm!(d-m-l-1)!(l+1)!(n-d)!(1-p_0)},\\ p_{11,b,e}= & {} -p_{11,a,e}e^{-\frac{(T-\mu )(l+1)}{\sigma }},\\ p_{12,e}= & {} \frac{(-1)^{l_1+l_2}n!e^{-\frac{(T-\mu )(n-d+l_2)}{\sigma }}}{2^nm!(j-m-l_1-1)!l_1!(d-j-l_2)!l_2!(n-d)!(m+l_1+1)(1-p_0)},\\ p_{13,a}= & {} \frac{(-1)^{l_1+l_2}n!}{2^nj!(m-1-j-l_1)!(l_1+1)!(k-m-2-l_2)!l_2!(n-k)!(n-k+l_2+1)m(1-p_0)},\\ p_{13,b}= & {} -p_{13,a}e^{-\frac{(T-\mu )m}{\sigma }},\\ p_{13,c}= & {} \frac{(-1)^{l_1+l_2+1}n!e^{-\frac{(T-\mu )(n-k+l_2+1)}{\sigma }}}{2^nj!(m-1-j-l_1)!(l_1+1)!(k-m-l_2-1)!l_2!(n-k)!(n-k+l_2+1)(1-p_0)},\\ p_{13,d}= & {} -p_{13,c}e^{-\frac{(T-\mu )(k-m-l_2-1)}{\sigma }},\\ p_{13,e}= & {} \frac{(-1)^{l_1+l_2+1}n!}{2^nj!(m-1-j-l_1)!(l_1+1)!(k-m-2-l_2)!l_2!(n-k)!(n-k+l_2+1)(m+l_1+1)(1-p_0)},\\ p_{13,f}= & {} -p_{13,e}e^{-\frac{(T-\mu )(m+l_1+1)}{\sigma }},\\ p_{13,g}= & {} \frac{(-1)^{l_1+l_2}n!e^{-\frac{n-k+l_2+1}{\sigma }}}{2^nj!(m-1-j-l_1)!(l_1+1)!(k-m-2-l_2)!l_2!(n-k)!(n-k+l_2+1)(k-m-l_2+l_1)(1-p_0)},\\ p_{13,h}= & {} -p_{13,g}e^{-\frac{(T-\mu )(k-m-l_2+l_1)}{\sigma }},\\ p_{13,a}^*= & {} \frac{(-1)^{l_1+k-m-l_2}n!}{2^n j!(m-j-l_1)!l_1!(k-m-2-l_2)!l_2!(n-k)!(m-l_2-1)(m+l_1)},\\ p_{13,b}^*= & {} -p_{13,a}^*e^{-\frac{(T-\mu )(m+l_1)}{\sigma }},\\ p_{13,c}^*= & {} \frac{(-1)^{l_1+k-m-1-l_2}n!e^{-\frac{(T-\mu )(m-l_2-1)}{\sigma }}}{2^n j!(m-j-l_1)!l_1!(k-m-2-l_2)!l_2!(n-k)!(m-l_2-1)(l_1+l_2+1)},\\ p_{13,d}^*= & {} -p_{13,c}^*e^{-\frac{(T-\mu )(l_1+l_2+1)}{\sigma }},\\ p_{14,a}= & {} \frac{(-1)^ln!}{2^nm!(k-m-2-l)!l!(n-k+l+1)m(1-p_0)},\\ p_{14,b}= & {} -p_{14,a}e^{-\frac{(T-\mu )m}{\sigma }},\\ p_{14,c}= & {} \frac{(-1)^{l+1}n!e^{-\frac{(T-\mu )(n-k+l+1)}{\sigma }}}{2^nm!(k-m-1-l)!l!(n-k+l+1)(1-p_0)},\\ p_{14,d}= & {} -p_{14,c}e^{-\frac{(T-\mu )(k-m-1-l)}{\sigma }},\\ p_{15,a}= & {} \frac{n!(-1)^{l_1+l_2}(m+l_1+1)^{-1}}{2^nm!(j-m-1-l_1)!l_1!(k-1-j-l_2)!l_2!(n-k)!(n-k+l_2+1)(1-p_0)},\\ p_{15,b}= & {} -p_{15,a}e^{-\frac{(T-\mu )(n-k+l_2+1)}{\sigma }},\\ p_{16,a}= & {} \frac{(-1)^{l_1+l_2}n!}{2^{l_2+d}m!(k-m-2-l_1)!l_1!(n-k-l_2)!l_2!(k-m-1-l_1+l_2)(m+l_1+1)(1-p_0)},\\ p_{16,b}= & {} \frac{(-1)^{l_1+l_2+1}n!}{2^{l_2+d}m!(k-m-2-l_1)!l_1!(n-k-l_2)!l_2!(k-m-1-l_1+l_2)(k+l_2)(1-p_0)},\\ q_1= & {} P(D=d),\\ q_2= & {} \frac{(-1)^le^{\frac{(T-\mu )(k+l)}{\sigma }}n!}{2^{k+l} (k-1)!l!(n-k-l)!(k+l)(1-q_0)},\\ q_{3,e}= & {} P(D=m),\\ q_{4,e}= & {} \frac{(-1)^l n!e^{\frac{(T-\mu )(m+l+1)}{\sigma }}}{2^{k+l} m! l!(m-1-l)!(m+l+1)(1-q_0)},\\ q_5= & {} \frac{(-1)^l\left( 1-\frac{1}{2}e^{-\frac{\mu -T}{\sigma }}\right) ^{n-d} \left( \frac{1}{2}e^{-\frac{(\mu -T)}{\sigma }}\right) ^d n!}{(l+m+1)m!l!(d-m-1-l)!(n-d)!(1-q_0)}\\ q_6= & {} \frac{(-1)^{l_1+l_2}e^{\frac{(T-\mu )(k+l_2)}{\sigma }} n!}{2^{k+l_2}(l_1+m+1)(k+l_2)m!l_1!(k-m-2-l_1)!l_2!(n-k-l_2)!(1-q_0)}\\ p_{4,a,o}= & {} \frac{(-1)^{l_1+d-m-1-l_2}n!e^{-\frac{(T-\mu )(m-l_2)}{\sigma }}}{ (1-p_0)2^n(l_1+l_2+1)l_1!l_2!j!(n-d)!(m-j-l_1)!(d-m-1-l_2)!},\\ p_{4,b,o}= & {} -p_{4,a,o}e^{-\frac{(T-\mu )(l_1+l_2+1)}{\sigma }},\\ p_{5,o}= & {} \frac{(-1)^{l_1+l_2}n!e^{-\frac{(T-\mu )(n-d+l_2)}{\sigma }}}{2^nm!(j-m-l_1-1)!l_1!(d-j-l_2)!l_2!(n-d)!(m+l_1+1)(1-p_0)},\\ p_{6,a,o}= & {} \frac{(-1)^{l_1+l_2}n!}{ 2^nj!(m-j-l_1)!l_1!(k-m-2-l_2)!l_2!(n-k)!(n-k+l_2+1)(m+l_1+1)(1-p_0)},\\ p_{6,b,o}= & {} -p_{6,a,o}e^{-\frac{(T-\mu )(m+l_1+1)}{\sigma }},\\ p_{6,c,o}= & {} \frac{(-1)^{l_1+l_2+1}n!e^{-\frac{(T-\mu )(n-d+l_2+1)}{\sigma }}}{ (1-p_0)2^nj!(m-j-l_1)!l_1!(k-m-2-l_2)!l_2!(n-k)!(n-k+l_2+1)(k-m-l_2+l_1-1)},\\ p_{6,d,o}= & {} -p_{6,c,o}e^{-\frac{(T-\mu )(d-m+l_1-l_2-1)}{\sigma }},\\ p_{7,a,o}= & {} \frac{(-1)^{l_1+l_2}n!}{ 2^nm!(j-m-1-l_1)!l_1!(k-1-j-l_2)!l_2!(n-k)!(m+l_1+1)(n-k+l_2+1)(1-p_0)},\\ p_{7,b,o}= & {} -p_{7,a,o}e^{-\frac{(T-\mu )(n-k+l_2+1)}{\sigma }},\\ p_{8,a,o}= & {} \frac{(-1)^{k-m-2-l_1+l_2}n!}{ 2^{k+l_2}m!(k-m-2-l_1)!l_1!(n-k-l_2)!l_2!(l_1+l_2+1)(k-l_1-1)(1-p_0)},\\ p_{8,b,o}= & {} \frac{(-1)^{k-m-1-l_1+l_2}n!}{ 2^{k+l_2}m!(k-m-2-l_1)!l_1!(n-k-l_2)!l_2!(l_1+l_2+1)(k+l_2)(1-p_0)},\\ \end{aligned}$$
$$\begin{aligned}&q_{3,o}=P(D=m+1),\\&p_{16}^*=\frac{(-1)^{l_1+l_2}n!}{m!(k-m-2-l_1)!l_1!(n-k-l_2)!l_2!(m+1+l_1)(k+l_2)(1-p_0)},\\&Z_{p1}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{d}\right) +N\Gamma \left( d-j,\frac{\sigma }{d}\right) +\frac{(T-\mu )(d-l)}{d},\\&Z_{p1}^{(2)}\overset{d}{=}\log \left( \frac{n}{2d}\right) Z_{p1}^{(1)}+T,\\&Z_{p2,a}^{(1)}\overset{d}{=}\Gamma _{A}\left( j,\frac{\sigma }{k}\right) +N\Gamma _{A}\left( k-1-j,\frac{\sigma }{k}\right) +\frac{(k-l-1)\sigma }{(n-k+l+1)k}E_1,\\&Z_{p2,a}^{(2)}\overset{d}{=}\log \left( \frac{n}{2k}\right) \left[ \Gamma _{A}\left( j,\frac{\sigma }{k}\right) +N\Gamma _{A}\left( k-1-j,\frac{\sigma }{k}\right) \right] +\left[ \log \left( \frac{n}{2k}\right) \frac{k-l-1}{n-k+l+1}+\frac{k}{n-k+l+1}\right] \frac{\sigma }{k} E_1+\mu ,\\&Z_{p2,a}^{(3)}\overset{d}{=}\frac{\sigma }{n-k+l+1}E_1+\mu ,\\&Z_{p2,b}^{(1)}\overset{d}{=}\Gamma _{A}\left( j,\frac{\sigma }{k}\right) +N\Gamma _{A}\left( k-1-j,\frac{\sigma }{k}\right) +\frac{(k-l-1)\sigma }{(n-k+l+1)k}E_2+\frac{(T-\mu )(k-l-1)}{k},\\&Z_{p2,b}^{(2)}\overset{d}{=}\log \left( \frac{n}{2k}\right) \left[ \Gamma _{A}\left( j,\frac{\sigma }{k}\right) +N\Gamma _{A}\left( k-1-j,\frac{\sigma }{k}\right) +\frac{(T-\mu )(k-l-1)}{k}\right] \\&+\left[ \log \left( \frac{n}{2k}\right) \frac{k-l-1}{n-k+l+1}+\frac{k}{n-k+l+1}\right] \frac{\sigma }{k} E_1+T,\\&Z_{p2,b}^{(3)}\overset{d}{=} \frac{\sigma }{n-k+l+1}E_1+T\\&Z_{p3}^{(1)}\overset{d}{=}\Gamma \left( k-1,\frac{\sigma }{k}\right) ,\\&Z_{p3}^{(2)}\overset{d}{=}\log \left( \frac{n}{2k}\right) Z_{p3}^{(1)}+NE\left( \frac{\sigma }{k+l}\right) +\mu ,\\&Z_{p3}^{(3)}\overset{d}{=}NE\left( \frac{\sigma }{k+l}\right) +\mu ,\\&Z_{p4,a,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{m}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m}\right) -\frac{\sigma }{m}E+(T-\mu ),\\&Z_{p4,a,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E+\frac{T+\mu }{2},\\&Z_{p4,b,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{m}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m}\right) -\frac{\sigma }{m}E+\frac{(T-\mu )(m-l-1)}{m},\\&Z_{p4,b,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E+T,\\&Z_{p5,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{m}\right) +\frac{\sigma }{m}E+(T-\mu ),\\&Z_{p5,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E+\frac{T+\mu }{2},\\&Z_{p6,a,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{m+1}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m+1}\right) -\frac{\sigma }{m+1}E_1+\frac{\sigma }{m+1}E_2,\\&Z_{p6,a,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E_{1}+\frac{\sigma }{n}E_{2}+\mu ,\\&Z_{p6,b,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{m+1}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m+1}\right) -\frac{\sigma }{m+1}E_{1}+\frac{\sigma }{m+1}E_{2}+\frac{m(T-\mu )}{m+1},\\&Z_{p6,b,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E_{1}+\frac{\sigma }{n}E_{2}+\frac{T+\mu }{2},\\&Z_{p6,c,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{m+1}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m+1}\right) -\frac{\sigma }{m+1}E_{1}+\frac{\sigma (m-l-1)}{(m+l+1)(m+1)}E_{2},\\&Z_{p6,c,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E_{1}+\frac{\sigma }{m+l+1}E_{2}+\mu ,\\&Z_{p6,d,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{m+1}\right) -\frac{\sigma }{m+1}E_{1}+\frac{\sigma (m-l-1)}{(m+l+1)(m+1)}E_{2}+\frac{(T-\mu )(m-l-1)}{m+1},\\&Z_{p6,d,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l+1)}E_{1}+\frac{\sigma }{m+l+1}E_{2}+T,\\&Z_{p7,a,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{m+1}\right) +\frac{\sigma }{k}E_{1}+\frac{\sigma }{m+1}E_{2},\\&Z_{p7,a,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_{1}+\frac{\sigma }{n}E_{2}+\mu ,\\&Z_{p7,b,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{m+1}\right) +\frac{\sigma }{m+1}E_1+\frac{\sigma }{m+1}E_2+\frac{m(T-\mu )}{m+1},\\&Z_{p7,b,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{n}E_2+\frac{(T+\mu )}{2},\\&Z_{p9,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E,\\&Z_{p9,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E+NE\left( \frac{\sigma }{m+l+1}\right) +\mu ,\\&Z_{p10,a,e}^{(1)}\overset{d}{=}\Gamma \left( d-m-1+j,\frac{\sigma }{d}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{d}\right) +\frac{\sigma }{d}E_1-\frac{\sigma }{d}E_2+\frac{(T-\mu )(m-1-l_2)}{d},\\&Z_{p10,a,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_2+1)}E_1+\frac{\sigma }{2(l_1+1)}E_2+\mu ,\\&Z_{p10,b,e}^{(1)}\overset{d}{=}\Gamma \left( d-m-1+j,\frac{\sigma }{d}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{d}\right) +\frac{\sigma }{d}E_1-\frac{\sigma }{d}E_2 +\frac{(T-\mu )m}{d},\\&Z_{p10,b,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_2+1)}E_1+\frac{\sigma }{2(l_1+1)}E_2+\frac{T+\mu }{2},\\&Z_{p10,c,e}^{(1)}\overset{d}{=}\Gamma \left( d-m-1+j,\frac{\sigma }{d}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{d}\right) +\frac{(l_2-l_1)\sigma }{(l_1+l_2+2)d}E_1-\frac{\sigma }{d}E_2+\frac{(T-\mu )(m-l_2-1)}{d},\\&Z_{p10,c,e}^{(2)}\overset{d}{=}\frac{\sigma }{l_1+l_2+2}E_1+\frac{\sigma }{2(l_1+1)}E_2+\mu ,\\&Z_{p10,d,e}^{(1)}\overset{d}{=}\Gamma \left( d-m-1+j,\frac{\sigma }{d}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{d}\right) +\frac{(l_2-l_1)\sigma }{(l_1+l_2+2)d}E_1-\frac{\sigma }{d}E_2 +\frac{(T-\mu )(m-l_1-1)}{d},\\&Z_{p10,d,e}^{(2)}\overset{d}{=}\frac{\sigma }{l_1+l_2+2}E_1+\frac{\sigma }{2(l_1+1)}E_2+T,\\&Z_{p11,a,e}^{(1)}\overset{d}{=}\Gamma \left( d-2,\frac{\sigma }{d}\right) +\frac{\sigma }{d}E_1+\frac{\sigma }{d}E_2+\frac{(T-\mu )(m-l-1)}{d},\\&Z_{p11,a,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{2(l+1)}E_2+\mu ,\\&Z_{p11,b,e}^{(1)}\overset{d}{=}\Gamma \left( d-2,\frac{\sigma }{d}\right) +\frac{\sigma }{d}E_1+\frac{\sigma }{d}E_2+\frac{(T-\mu )m}{d},\\&Z_{p11,b,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{2(l+1)}E_2+\frac{T+\mu }{2},\\&Z_{p12,e}^{(1)}\overset{d}{=}\Gamma \left( d-j+m-1,\frac{\sigma }{d}\right) +N\Gamma \left( j-m-1,\frac{\sigma }{d}\right) +\frac{\sigma }{d}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)d}E_2 +\frac{(T-\mu )(n-d+l_2)}{d}, \end{aligned}$$
$$\begin{aligned}&Z_{p12,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1-\frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{p13,a}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2,\\&Z_{p13,a}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{n}E_2+\mu ,\\&Z_{p13,b}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2+\frac{m(T-\mu )}{k},\\&Z_{p13,b}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{n}E_2+\frac{T+\mu }{2},\\&Z_{p13,c}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2+\frac{(n-k+l_2+1)(T-\mu )}{k},\\&Z_{p13,c}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{2(k-m-l_2-1)}E_2+\mu ,\\&Z_{p13,d}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2+\frac{m(T-\mu )}{k},\\&Z_{p13,d}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{2(k-m-l_2-1)}E_2+\frac{T+\mu }{2},\\&Z_{p13,e}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2,\\&Z_{p13,e}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{p13,f}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2 +\frac{(m-l_1-1)(T-\mu )}{k},\\&Z_{p13,f}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{m+l_1+1}E_2+T,\\&Z_{p13,g}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{(k-m-l_1-l_2-2)\sigma }{(k-m-l_2+l_1)k}E_2\\&\qquad \quad +\frac{(n-k+l_2+1)(T-\mu )}{k},\\&Z_{p13,g}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{k-m-l_2+l_1}E_2+\mu ,\\&Z_{p13,h}^{(1)}\overset{d}{=}\Gamma \left( k-m-1+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) -\frac{\sigma }{k}E_1+\frac{(k-m-l_1-l_2-2)\sigma }{(k-m-l_2+l_1)k}E_2\\&\qquad \quad +\frac{(m-l_1-1)(T-\mu )}{k},\\&Z_{p13,h}^{(2)}\overset{d}{=}\frac{\sigma }{2(l_1+1)}E_1+\frac{\sigma }{k-m-l_2+l_1}E_2+T,\\&Z_{p13,a}^*\overset{d}{=}\Gamma \left( k-1-m+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-j,\frac{\sigma }{k}\right) +E\left( \frac{(m-l_1)\sigma }{(m+l_1)k}\right) ,\\&Z_{p13,b}^*\overset{d}{=}\Gamma \left( k-1-m+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-j,\frac{\sigma }{k}\right) +E\left( \frac{(m-l_1)\sigma }{(m+l_1)k}\right) +\frac{(T-\mu )(m-l_1)}{k},\\&Z_{p13,c}^*\overset{d}{=}\Gamma \left( k-1-m+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-j,\frac{\sigma }{k}\right) +E\left( \frac{(l_2-l_1+1)\sigma }{(l_1+l_2+1)k}\right) +\frac{(T-\mu )(m-l_2-1)}{k},\\&Z_{p13,d}^*\overset{d}{=}\Gamma \left( k-1-m+j,\frac{\sigma }{k}\right) +N\Gamma \left( m-j,\frac{\sigma }{k}\right) +E\left( \frac{(l_2-l_1+1)\sigma }{(l_1+l_2+1)k}\right) +\frac{(T-\mu )(m-l_1)}{k},\\&Z_{p14,a}^{(1)}\overset{d}{=}\Gamma \left( k-2,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1 +\frac{\sigma }{k}E_2,\\&Z_{p14,a}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{n}E_2+\mu ,\\&Z_{p14,b}^{(1)}\overset{d}{=}\Gamma \left( k-2,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1 +\frac{\sigma }{k}E_2+\frac{m(T-\mu )}{k},\\&Z_{p14,b}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{n}E_2+\frac{T+\mu }{2},\\&Z_{p14,c}^{(1)}\overset{d}{=}\Gamma \left( k-2,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1 +\frac{\sigma }{k}E_2+\frac{(T-\mu )(n-k+l+1)}{k},\\&Z_{p14,c}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{2(k-m-l-1)}E_2+\mu ,\\&Z_{p14,d}^{(1)}\overset{d}{=}\Gamma \left( k-2,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1 +\frac{\sigma }{k}E_2+\frac{(T-\mu )m}{k},\\&Z_{p14,d}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{2(k-m-l-1)}E_2+\frac{T+\mu }{2},\\&Z_{p15,a}^{(1)}\overset{d}{=}\Gamma \left( m-1+k-j,\frac{\sigma }{k}\right) + N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1 +\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2,\\&Z_{p15,a}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1 -\frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{p15,b}^{(1)}\overset{d}{=}\Gamma \left( m-1+k-j,\frac{\sigma }{k}\right) + N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2+\frac{(T-\mu )(n-k+l_2+1)}{k},\\&Z_{p15,b}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1 -\frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{p16,a}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) + N\Gamma \left( d-m-2,\frac{\sigma }{k}\right) +NE\left( \frac{(m-l_1-1)\sigma }{(k-m-1-l_1+l_2)k}\right) +\frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2,\\&Z_{p16,a}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1 -\frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{p16,b}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) + N\Gamma \left( d-m-2,\frac{\sigma }{k}\right) +NE\left( \frac{(m-l_1-1)\sigma }{(k-m-1-l_1+l_2)k}\right) +\frac{\sigma }{k}E,\\&Z_{p16,b}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E +NE\left( \frac{\sigma }{k+l_2}\right) +\mu ,\\&Z_{q1}^{(1)}\overset{d}{=}\Gamma \left( d,\frac{\sigma }{d}\right) ,\\&Z_{q1}^{(2)}\overset{d}{=}\log \left( \frac{n}{2d}\right) Z_{q1}^{(1)}+T,\\&Z_{q2,a}^{(1)}\overset{d}{=}\Gamma \left( k-1,\frac{\sigma }{k}\right) ,\\&Z_{q2,a}^{(2)}\overset{d}{=}\log \left( \frac{n}{2k}\right) Z_{q2,a}^{(1)}+NE\left( \frac{\sigma }{k+l}\right) +T,\\&Z_{q2,b}^{(1)}\overset{d}{=}\Gamma \left( k-1,\frac{\sigma }{k}\right) ,\\&Z_{q2,b}^{(2)}\overset{d}{=}NE\left( \frac{\sigma }{k+l}\right) +T,\\&Z_{q3,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{m}\right) +\frac{\sigma }{m}E,\\&Z_{q3,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E+T,\\&Z_{q4,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E,\\&Z_{q4,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E+NE\left( \frac{\sigma }{m+l+1}\right) +T,\\&Z_{q5,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{d}\right) +N\Gamma \left( d-m-1,\frac{\sigma }{d}\right) + \frac{\sigma }{d}E_1+\frac{(m-l-1)\sigma }{(m+l+1)d}E_2,\\&Z_{q5,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1-\frac{\sigma }{m+l+1}E_2+T,\\&Z_{q6,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) +N\Gamma \left( k-m-2,\frac{\sigma }{k}\right) + \frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{(m+l_1+1)k}E_2,\\&Z_{q6,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1-\frac{\sigma }{m+l_1+1}E_2+NE\left( \frac{\sigma }{k+l_2}\right) +T,\\&Z_{p4,a,o}^{(1)}\overset{d}{=}\Gamma \left( j+d-m-1,\frac{\sigma }{d}\right) + N\Gamma \left( m-j,\frac{\sigma }{d}\right) +\frac{(l_2-l_1)\sigma }{(l_2+l_1+1)d} E +\frac{(T-\mu )(m-l_2)}{d},\\&Z_{p4,a,o}^{(2)}\overset{d}{=}\frac{\sigma }{l_2+l_1+1}E+\mu ,\\&Z_{p4,b,o}^{(1)}\overset{d}{=}\Gamma \left( j+d-m-1,\frac{\sigma }{d}\right) + N\Gamma \left( m-j,\frac{\sigma }{d}\right) +\frac{(l_2-l_1)\sigma }{(l_2+l_1+1)d} E+\frac{(T-\mu )(m-l_1)}{d},\\&Z_{p4,b,o}^{(2)}\overset{d}{=}\frac{\sigma }{l_2+l_1+1}E+T,\\&Z_{p5,o}^{(1)}\overset{d}{=}\Gamma \left( d-j+m,\frac{\sigma }{d}\right) + N\Gamma \left( j-m-1,\frac{\sigma }{d}\right) +\frac{(m-l_1)\sigma }{(m+l_1+1)d} E +\frac{(T-\mu )(n-d+l_2)}{d},\\&Z_{p5,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{p6,a,o}^{(1)}\overset{d}{=}\Gamma \left( j+k-m-1,\frac{\sigma }{k}\right) + N\Gamma \left( m-j,\frac{\sigma }{k}\right) +\frac{(m-l_1)\sigma }{(m+l_1+1)k} E,\\&Z_{p6,a,o}^{(2)}\overset{d}{=}\frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{p6,b,o}^{(1)}\overset{d}{=}\Gamma \left( j+k-m-1,\frac{\sigma }{k}\right) + N\Gamma \left( m-j,\frac{\sigma }{k}\right) +\frac{(m-l_1)\sigma }{(m+l_1+1)k} E+\frac{(m-l_1)(T-\mu )}{k},\\&Z_{p6,b,o}^{(2)}\overset{d}{=}\frac{\sigma }{m+l_1+1}E+T,\\&Z_{p6,c,o}^{(1)}\overset{d}{=}\Gamma \left( j+k-m-1,\frac{\sigma }{k}\right) + N\Gamma \left( m-j,\frac{\sigma }{k}\right) ++\frac{(k-m-l_1-l_2-2)\sigma }{(k-m+l_1-l_2-1)k} E+\frac{(n-k+l_2+1)(T-\mu )}{k},\\&Z_{p6,c,o}^{(2)}\overset{d}{=}\frac{\sigma }{k-m-l_2+l_1-1}E+\mu ,\\&Z_{p6,d,o}^{(1)}\overset{d}{=}\Gamma \left( j+k-m-1,\frac{\sigma }{k}\right) + N\Gamma \left( m-j,\frac{\sigma }{k}\right) +\frac{(k-m-l_1-l_2-2)\sigma }{(k-m+l_1-l_2-1)k} E+\frac{(m-l_1)(T-\mu )}{k},\\&Z_{p6,d,o}^{(2)}\overset{d}{=}\frac{\sigma }{k-m+l_1-l_2-1}E+T,\\&Z_{p7,a,o}^{(1)}\overset{d}{=}\Gamma \left( m+k-j,\frac{\sigma }{k}\right) +N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{(m-l_1)\sigma }{(m+l_1+1)k}E,\\&Z_{p7,a,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{p7,b,o}^{(1)}\overset{d}{=}\Gamma \left( m+k-j,\frac{\sigma }{k}\right) +N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{(m-l_1)\sigma }{(m+l_1+1)k}E+\frac{(n-k+1+l_2)(T-\mu )}{k},\\&Z_{p7,b,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{p8,a,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{k}\right) +N\Gamma \left( k-m-2,\frac{\sigma }{k}\right) +NE\left( \frac{n-k+l_1+1}{(l_1+l_2+1)k}\right) +\frac{n+1+l_1-k}{(k-l_1-1)k}E,\\&Z_{p8,a,o}^{(2)}\overset{d}{=}-\frac{\sigma }{k-l_1-1}E+\mu ,\\&Z_{p8,b,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{k}\right) +N\Gamma \left( k-m-2,\frac{\sigma }{k}\right) +NE\left( \frac{n-k+l_1+1}{(l_1+l_2+1)k}\right) ,\\&Z_{p8,b,o}^{(2)}\overset{d}{=}NE\left( \frac{\sigma }{k+l_2}\right) +\mu ,\\&Z_{q3,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{m+1}\right) +\frac{\sigma }{m+1}E,\\&Z_{q3,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+1}E+T,\\&Z_{q5,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{d}\right) +N\Gamma \left( d-m-1,\frac{\sigma }{d}\right) + \frac{(m-l)\sigma }{(m+l+1)d}E,\\&Z_{q5,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l+1}E+T,\\&Z_{q6,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{k}\right) +N\Gamma \left( k-m-2,\frac{\sigma }{k}\right) + \frac{(m-l_1)\sigma }{(m+l_1+1)k}E,\\&Z_{q6,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l_1+1}E+NE\left( \frac{\sigma }{k+l_2}\right) +T,\\&Z_{p16}^*\overset{d}{=}\Gamma \left( m,\frac{\sigma }{k}\right) +N\Gamma \left( k-2-m,\frac{\sigma }{m}\right) +E\left( \frac{(m-l_1-1)\sigma }{k(m+l_1+1)}\right) . \end{aligned}$$

List of notation used in Lemma 7.3

$$\begin{aligned}&p_{a1}=\frac{(-1)^l n!}{2^nj!(k-1-j-l)!l!(n-k)!(n-k+l+1)},\\&p_{a2}=\frac{(-1)^l n!}{2^{k+l}(k-1)!l!(n-k-l)!(k+l)},\\&p_{b1,e}=\frac{(-1)^l n!}{2^nj!(k-1-j-l)!l!(n-k)!(k+l+1)},\\&p_{b2}=\frac{(-1)^l n!}{2^{k+l}(k-1)!(n-k-l)!l!(k+l)},\\&p_{b1,o}=\frac{(-1)^l n!}{2^nj!(k-1-j-l)!l!(n-k)!(k+l)},\\&p_{c1}=\frac{(-1)^l n!}{2^nj!(m-j-1-l)!l!(n-k+1)!(n-k+l+2)},\\&p_{c2}=\frac{n!}{2^n m!(n-k+1)!},\\&p_{c3}=\frac{(-1)^ln! }{2^{k+l}m!(n-k-l)!l! (l+m+1)},\\&p_{d1,e}=\frac{(-1)^{l_1+l_2}n!}{2^nj!(m-1-j-l_1)!l_1!(k-2-m-l_2)!l_2!(n-k)! m(m+l_1+1)(n-k+l_2+1)},\\&p_{d1,o}=\frac{(-1)^{l_1+l_2}n!}{2^nj!(m-j-l_1)!l_1!(k-2-m-l_2)!l_2!(n-k)! (m+l_1+1)(n-k+l_2+1)},\\&p_{d2}=\frac{(-1)^{l}n!}{2^nm!(k-2-m-l)!l!(n-k)!(n-r+l+1) m},\\&p_{d3}=\frac{(-1)^{l_1+l_2}n!}{2^nm!(j-1-m-l_1)!l_1!(k-1-j-l_2)!l_2!(n-k)! (l_1+m+1)(n-k+1+l_2)},\\&p_{d4}=\frac{(-1)^{l_1+l_2}n!}{2^{k+l_1}m!(k-2-m-l_2)!l_2!(n-k-l_1)!l_1!(k+l_1) (m+l_2+1)},\\&Z_{a1}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{k}\right) +N\Gamma \left( k-1-j,\frac{\sigma }{k}\right) +\frac{k-1-l}{k(n-r+l+1)}E,\\&Z_{a1}^{(2)}\overset{d}{=} \ln \left( \frac{n}{2k}\right) Z_{a1}^{(1)}+\frac{\sigma }{(n-r+l+1)}E+\mu ,\\&Z_{a2}^{(1)}\overset{d}{=}\Gamma \left( k-1,\frac{\sigma }{k}\right) ,\\&Z_{a2}^{(2)}\overset{d}{=}\ln \left( \frac{n}{2k}\right) Z_{a2}^{(1)}+NE\left( \frac{1}{l+k}\right) +\mu ,\\&Z_{b1,e}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{k}\right) +N\Gamma \left( k-1-j,\frac{\sigma }{k}\right) +\frac{(k-l-1)\sigma }{k(k+l+1)}E,\\&Z_{b1,e}^{(2)}\overset{d}{=}\frac{\sigma }{(k+l+1)}E+\mu ,\\&Z_{b1,o}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{k}\right) +N\Gamma \left( k-1-j,\frac{\sigma }{k}\right) +\frac{(k-1-l)\sigma }{k(k+l)}E,\\&Z_{b1,o}^{(2)}\overset{d}{=}\frac{\sigma }{k+l}E+\mu ,\\&Z_{b2}^{(1)}\overset{d}{=}\Gamma \left( k-1,\frac{\sigma }{k}\right) ,\\&Z_{b2}^{(2)}\overset{d}{=}NE\left( \frac{\sigma }{k+l}\right) +\mu ,\\&Z_{c1}^{(1)}\overset{d}{=}\Gamma \left( j,\frac{\sigma }{k}\right) +N\Gamma \left( m-j-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{(n-k-l)\sigma }{k(n-k+l+2)}E_2,\\&Z_{c1}^{(2)}\overset{d}{=}\frac{\sigma }{n}E_1+\frac{1}{n-k+l+2}E_2+\mu ,\\&Z_{c2}^{(1)}\overset{d}{=} \Gamma \left( m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2,\\&Z_{c2}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1+\frac{\sigma }{n}E_2+\mu ,\\&Z_{c3}^{(1)}\overset{d}{=} \Gamma \left( m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E,\\&Z_{c3}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E+NE\left( \frac{\sigma }{m+l+1}\right) +\mu ,\\&Z_{d1,e}^{(1)}\overset{d}{=} \Gamma \left( k+j-m-1,\frac{\sigma }{k}\right) +N\Gamma \left( m-1-j,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+ \frac{(m-l_1-1)\sigma }{k(m+l_1+1)}E_2,\\&Z_{d1,e}^{(2)}\overset{d}{=} \frac{\sigma }{n}E_1+ \frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{d1,o}^{(1)}\overset{d}{=} \Gamma \left( k+j-m-1,\frac{\sigma }{k}\right) +N\Gamma \left( m-j,\frac{\sigma }{k}\right) + \frac{(m-l_1)\sigma }{k(m+l_1+1)}E_2,\\&Z_{d1,o}^{(2)}\overset{d}{=} \frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{d2}^{(1)}\overset{d}{=} \Gamma \left( k-2,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{\sigma }{k}E_2,\\&Z_{d2}^{(2)}\overset{d}{=} -\frac{\sigma }{n}E_1+\frac{\sigma }{n}E_2+\mu ,\\&Z_{d3,e}^{(1)}\overset{d}{=}\Gamma \left( k+m-j-1,\frac{\sigma }{k}\right) +N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{(m-l_1-1)\sigma }{k(m+l_1+1)}E_2,\\&Z_{d3,e}^{(2)}\overset{d}{=}-\frac{\sigma }{n}E_1- \frac{\sigma }{m+l_1+1}E_2+\mu ,\\&Z_{d3,o}^{(1)}\overset{d}{=}\Gamma \left( k+m-j,\frac{\sigma }{k}\right) +N\Gamma \left( j-m-1,\frac{\sigma }{k}\right) +\frac{(m-l_1)\sigma }{k(m+l_1+1)}E,\\&Z_{d3,o}^{(2)}\overset{d}{=}-\frac{\sigma }{m+l_1+1}E+\mu ,\\&Z_{d4,e}^{(1)}\overset{d}{=}\Gamma \left( m-1,\frac{\sigma }{k}\right) +N\Gamma \left( k-2-m,\frac{\sigma }{k}\right) +\frac{\sigma }{k}E_1+\frac{(m-l_2-1)\sigma }{k(m+l_2+1)}E_2,\\&Z_{d4,e}^{(2)}\overset{d}{=} -\frac{\sigma }{n}E_1-\frac{\sigma }{m+l_2+1}E_2 +NE\left( \frac{\sigma }{k+l_1}\right) +\mu ,\\&Z_{d4,o}^{(1)}\overset{d}{=}\Gamma \left( m,\frac{\sigma }{k}\right) +N\Gamma \left( k-2-m,\frac{\sigma }{k}\right) +\frac{(m-l_2)\sigma }{k(m+l_2+1)}E,\\&Z_{d4,o}^{(2)}\overset{d}{=} -\frac{\sigma }{m+l_2+1}E +NE\left( \frac{\sigma }{k+l_1}\right) +\mu . \end{aligned}$$

Proof of Lemma 3.4 Here, we will only give the proof for the case when \(k>m+1\), and all other cases can all be proved similarly. We have

$$\begin{aligned} E\left( e^{t{\hat{\sigma }}+s{\hat{\mu }}}, D=k|D>0\right)= & {} \sum \limits _{j=0}^{m-1}E\left( e^{t{\hat{\sigma }}+s{\hat{\mu }}}, D=k, J=j|D>0\right) \\&+\,E\left( e^{t{\hat{\sigma }}+s{\hat{\mu }}}, D=k, J=m|D>0\right) \\&+\,\sum \limits _{j=m+1}^{k-1}E\left( e^{t{\hat{\sigma }}+s\hat{\mu }}, D=k, J=j|D>0\right) \\&+\,E\left( e^{t{\hat{\sigma }}+s{\hat{\mu }}}, D=k, J=k|D>0\right) , \end{aligned}$$

where J is the number of X’s less than \(\mu \) as defined earlier in Sect. 3. As the derivation of these four expectations are quite similar, we only derive the last one and omit others for the sake of brevity. In this case, we only need to focus on three order statistics \(X_{m:n}\), \(X_{m+1:n}\) and \(X_{k:n}\), and then consider that there exists \(m-1\) i.i.d. observations less than \(X_{m:n}\), and \(k-m-2\) i.i.d. observations are between \(X_{m+1:n}\) and \(X_{k:n}\). We then have

$$\begin{aligned} E\left( e^{t{\hat{\sigma }}+s\hat{\mu }}, D=k, J=k|D>0\right)= & {} \frac{n!}{(m-1)!(k-m-2)!(n-k)!(1-p_0)} \int _{-\infty }^\mu \int _{-\infty }^{X_{m+1:n}}\int _{X_{m+1:n}}^{\mu }\\&\,\times \, \left[ \int _{-\infty }^{X_{m:n}} e^{-tx}\frac{1}{2\sigma }e^{\frac{x-\mu }{\sigma }}\right] ^{m-1} \left[ \int _{X_{m+1:n}}^{X_{k:n}} e^{tx}\frac{1}{2\sigma }e^{\frac{x-\mu }{\sigma }}\right] ^{k-m-2}\\&\,\times \, e^{t(n-k+1)x_{k:n}} \frac{1}{2\sigma }e^{\frac{x_{k:n}-\mu }{\sigma }}\left( 1-\frac{1}{2}e^{\frac{x_{k:n}-\mu }{\sigma }}\right) ^{n-k}dx_{k:n}\\&\,\times \, e^{\left( \frac{s}{2}-t\right) x_{m:n}}\frac{1}{2\sigma }e^{\frac{x_{m:n}-\mu }{\sigma }} dx_{m:n}\\&\,\times \, e^{\left( \frac{s}{2}+t\right) x_{m+1:n}}\frac{1}{2\sigma }e^{\frac{x_{m+1:n}-\mu }{\sigma }}dx_{m+1:n}\\= & {} \sum \limits _{l_1=0}^{k-m-2}\sum \limits _{l_2=0}^{n-k} \left[ p_{16,a}M_{Z_{p16,a}^{(1)},Z_{p16,a}^{(2)}}(t,s)+p_{16,b}M_{Z_{p16,b}^{(1)},Z_{p16,b}^{(2)}}(t,s)\right] . \end{aligned}$$

The last equation is obtained by using the binomial expansion.

In this case, if we are only interested in the marginal distribution of \({\hat{\sigma }}\), we can then focus only on \(X_{m+1:n}\) and \(X_{d:n}\) and consider that there exists m i.i.d. observations less then \(X_{m+1:n}\) and \(k-m-2\) i.i.d. observations between \(X_{m+1:n}\) and \(X_{k:n}\). We then have

$$\begin{aligned} E\left( e^{t{\hat{\sigma }}}, D=k, J=k|D>0\right)= & {} \frac{n!}{m!(k-m-2)!(n-k)!(1-p_0)} \int _{-\infty }^\mu \int _{-\infty }^{X_{k:n}}\\&\times \,\,\left[ \int _{-\infty }^{X_{m+1:n}} e^{-tx}\frac{1}{2\sigma }e^{\frac{x-\mu }{\sigma }}\right] ^{m} \left[ \int _{X_{m+1:n}}^{X_{k:n}} e^{tx}\frac{1}{2\sigma }e^{\frac{x-\mu }{\sigma }}\right] ^{k-m-2}\\&\times \, e^{tx_{m+1:n}}\frac{1}{2\sigma }e^{\frac{x_{m+1:n}-\mu }{\sigma }}dx_{m+1:n}\\&\times \,\, e^{t(n-k+1)x_{k:n}} \frac{1}{2\sigma }e^{\frac{x_{k:n}-\mu }{\sigma }}\left( 1-\frac{1}{2}e^{\frac{x_{k:n}-\mu }{\sigma }}\right) ^{n-k}dx_{k:n}\\= & {} \sum \limits _{l_1=0}^{k-m-2}\sum \limits _{l_2=0}^{n-k} p_{16}^* M_{Z^*_{p16}}(t), \end{aligned}$$

as required. \(\blacksquare \)

Rights and permissions

Reprints and permissions

Copyright information

© 2018 Springer International Publishing AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Su, F., Balakrishnan, N., Zhu, X. (2018). Exact Likelihood-Based Point and Interval Estimation for Lifetime Characteristics of Laplace Distribution Based on hybrid Type-I and Type-II Censored Data. In: Tez, M., von Rosen, D. (eds) Trends and Perspectives in Linear Statistical Inference . Contributions to Statistics. Springer, Cham. https://doi.org/10.1007/978-3-319-73241-1_13

Download citation

Publish with us

Policies and ethics