Skip to main content

Proportional Odds COM-Poisson Cure Rate Model with Gamma Frailty and Associated Inference and Application

  • Chapter
  • First Online:
Trends in Mathematical, Information and Data Sciences

Part of the book series: Studies in Systems, Decision and Control ((SSDC,volume 445))

  • 390 Accesses

Abstract

We introduce in this work a gamma frailty cure rate model for lifetime data by assuming the number of competing causes for the event of interest to follow the Conway-Maxwell-Poisson (COM-Poisson) distribution and the lifetimes of the non-cured individuals to follow a proportional odds model. The baseline distribution is taken to be either Weibull or log-logistic. Statistical inference is then developed under non-informative right censored data. We derive the maximum likelihood estimators (MLEs) with the use of Expectation Maximization (EM) method for all model parameters. The model discrimination among some well-known special cases, including Geometric, Poisson, and Bernoulli models, are discussed under both likelihood- and information-based criteria. An extensive Monte Carlo simulation study is carried out to examine the performance of the proposed model as well as all the inferential methods developed here. Finally, a cutaneous melanoma dataset is analyzed for illustrative purpose.

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 189.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 249.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 249.99
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

  1. Balakrishnan, N., Barui, S., Milienos, F.S.: Proportional hazards under Conway-Maxwell-Poisson cure rate model and associated inference. Stat. Meth. Med. Res. 26(5), 2055–2077 (2017)

    Article  MathSciNet  Google Scholar 

  2. Balakrishnan, N., Pal, S.: EM algorithm-based likelihood estimation for some cure rate models. J. Stat. Theory Pract. 6(4), 698–724 (2012)

    Article  MathSciNet  Google Scholar 

  3. Balakrishnan, N., Pal, S.: Lognormal lifetimes and likelihood-based inference for flexible cure rate models based on COM-Poisson family. Comput. Stat. Data Anal. 67, 41–67 (2013)

    Article  MathSciNet  Google Scholar 

  4. Balakrishnan, N., Pal, S.: COM-Poisson cure rate models and associated likelihood-based inference with exponential and Weibull lifetimes. In: Frenkel, I.B., Karagrigoriou, A., Lisnianski, A. (eds.) Applied Reliability Engineering and Risk Analysis: Probabilistic Models and Statistical Inference, pp. 308–348. Wiley, Chichester (2014)

    Google Scholar 

  5. Balakrishnan, N., Pal, S.: An EM algorithm for the estimation of parameters of a flexible cure rate model with generalized gamma lifetime and model discrimination using likelihood- and information-based methods. Comput. Stat. 30(1), 151–189 (2015)

    Article  MathSciNet  Google Scholar 

  6. Balakrishnan, N., Pal, S.: Expectation maximization-based likelihood inference for flexible cure rate models with Weibull lifetimes. Stat. Meth. Med. Res. 25(4), 1535–1563 (2016)

    Article  MathSciNet  Google Scholar 

  7. Berkson, J., Gage, R.P.: Survival curve for cancer patients following treatment. J. Am. Stat. Assoc. 47(259), 501–515 (1952)

    Article  Google Scholar 

  8. Boag, J.W.: Maximum likelihood estimates of the proportion of patients cured by cancer therapy. J. R. Stat. Soc. Ser. B 11(1), 15–53 (1949)

    MATH  Google Scholar 

  9. Conway, R.W., Maxwell, W.L.: A queuing model with state dependent service rates. J. Ind. Eng. 12(2), 132–136 (1962)

    Google Scholar 

  10. Ibrahim JG, Chen M-H, Sinha D.: Bayesian Survival Analysis. Springer, New York (2005)

    Google Scholar 

  11. Kadane, J.B., Shmueli, G., Minka, T.P., Borle, S., Boatwright, P.: Conjugate analysis of the Conway-Maxwell-Poisson distribution. Bayesian Anal. 1(2), 363–374 (2006)

    MathSciNet  MATH  Google Scholar 

  12. Kirkwood, J.M., Ibrahim, J.G., Sondak, V.K., Richards, J., Flaherty, L.E., Ernstoff, M.S., Smith, T.J., Rao, U., Steele, M., Blum, R.H.: High-and low-dose interferon alfa-2b in high-risk melanoma: first analysis of intergroup trial e1690/s9111/c9190. J. Clin. Oncol. 18(12), 2444–2458 (2000)

    Article  Google Scholar 

  13. McLachlan, G.J., Krishnan, T.: The EM Algorithm and Extensions. Wiley, Hoboken (2007)

    MATH  Google Scholar 

  14. Rodrigues, J., Cancho, V.G., de Castro, M., Louzada-Neto, F.: On the unification of long-term survival models. Stat. Probab. Lett. 79(6), 753–759 (2009)

    Article  MathSciNet  Google Scholar 

  15. Rodrigues, J., de Castro, M., Cancho, V.G., Balakrishnan, N.: COM-Poisson cure rate survival models and an application to a cutaneous melanoma data. J. Stat. Plan. Inference 139(10), 3605–3611 (2009)

    Article  MathSciNet  Google Scholar 

  16. Shmueli, G., Minka, T.P., Kadane, J.B., Borle, S., Boatwright, P.: A useful distribution for fitting discrete data: revival of the Conway-Maxwell-Poisson distribution. J. R. Stat. Soc. Ser. C 54(1), 127–142 (2005)

    Article  MathSciNet  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Narayanaswamy Balakrishnan .

Editor information

Editors and Affiliations

Appendices

Appendix

Expressions for Weibull baseline

In this case, we have the following expressions:

\( S_0(t)=e^{-{(\gamma _1 t)}^{1/\gamma _0}},f_0(t)=\frac{(\gamma _1 t)^{1/\gamma _0}}{\gamma _0 t}e^{-{(\gamma _1 t)}^{1/\gamma _0}}; \)

the derivatives of \(S_0\) are given by

$$\begin{aligned}&S_{0; \gamma _1} =\frac{(\gamma _1t_i)^{1/\gamma _0}\mathrm {log}(\gamma _1t_i)}{\gamma _0^2}S_0 ,~~ S_{0; \gamma _0} =-\frac{(\gamma _1t_i)^{1/\gamma _0}}{\gamma _0\gamma _1}S_0 ,~~ \\&S_{0; \gamma _0 \gamma _0} =S_{0; \gamma _0}\frac{[(\gamma _1t_i)^{1/\gamma _0}-1]\mathrm {log}(\gamma _1t_i)-2\gamma _0}{\gamma _0^2}, \\&S_{0; \gamma _1 \gamma _1} =S_{0; \gamma _1}\frac{[(\gamma _1t_i)^{1/\gamma _0}-1]\mathrm {log}(\gamma _1t_i)-\gamma _0}{\gamma _0^2} ,~~~ S_{0; \gamma _1\gamma _0} =S_{0; \gamma _1}\frac{1-\gamma _0-(\gamma _1t_i)^{1/\gamma _0}}{\gamma _0\gamma _1}; \end{aligned}$$

the derivatives of \(\log S_0\) are given by

$$\begin{aligned}&\frac{\partial \mathrm {log}S_0}{\partial \gamma _0}=\frac{(\gamma _1 t_i)^{1/\gamma _0}\mathrm {log}(\gamma _1 t_i)}{\gamma _0^2} ,~~ \frac{\partial \mathrm {log}S_0}{\partial \gamma _1}=-\frac{(\gamma _1 t_i)^{1/\gamma _0}}{\gamma _0\gamma _1} ,~~ \\&\frac{\partial ^2\mathrm {log} S_0}{\partial \gamma _0^2}=- \frac{\partial \mathrm {log}S_0}{\partial \gamma _0}\left( \frac{\mathrm {log}(\gamma _1 t_i)}{\gamma _0^2}+\frac{2}{\gamma _0} \right) , \\&\frac{\partial ^2\mathrm {log} S_0}{\partial \gamma _0\partial \gamma _1}=- \frac{\partial \mathrm {log}S_0}{\partial \gamma _1}\left( \frac{\mathrm {log}(\gamma _1 t_i)}{\gamma _0^2}+\frac{1}{\gamma _0}\right) ,~~ \frac{\partial ^2\mathrm {log} S_0}{\partial \gamma _1^2}=\frac{\partial \mathrm {log}S_0}{\partial \gamma _1}\left( \frac{1}{\gamma _0\gamma _1}-\frac{1}{\gamma _1}\right) ; \end{aligned}$$

the derivatives of \(\log f_0\) are given by

$$\begin{aligned}&\frac{\partial \mathrm {log} f_0}{\partial \gamma _0}=\frac{\partial \mathrm {log} S_0}{\partial \gamma _0}-\frac{1}{\gamma _0}\left( 1 +\frac{\mathrm {log}(\gamma _1t_i)}{\gamma _0}\right) ,~~ \frac{\partial \mathrm {log} f_0}{\partial \gamma _1}=\frac{1}{\gamma _0\gamma _1}+\frac{\partial \mathrm {log} S_0}{\partial \gamma _1} ,\\&\frac{\partial ^2\mathrm {log} f_0}{\partial \gamma _0^2}=\frac{\partial \mathrm {log} S_0^2}{\partial \gamma _0^2}+\frac{1}{\gamma _0^2}+\frac{2 \mathrm {log}(\gamma _1 t_i)}{\gamma _0^3} ,~~ \frac{\partial ^2\mathrm {log} f_0}{\partial \gamma _0\partial \gamma _1}=\frac{\partial \mathrm {log} S_0^2}{\partial \gamma _0 \partial \gamma _1}-\frac{1}{\gamma _0^2\gamma _1} ,~~ \\&\frac{\partial ^2\mathrm {log} f_0}{\partial \gamma _1 ^2}=\frac{\partial \mathrm {log}S_0^2}{\partial \gamma _1^2}-\frac{1}{\gamma _0\gamma _1^2}; \end{aligned}$$

the derivatives of S are given by

$$\begin{aligned}&\frac{\partial S}{ \partial \gamma _0}=\frac{\partial S_0}{ \partial \gamma _0} G_0 , ~~~~ \frac{\partial S}{ \partial \gamma _1} =\frac{\partial S_0}{ \partial \gamma _1} G_0 , ~~~~ \frac{\partial S}{\partial \alpha _l}= \frac{x_{il}F_0S}{G} , \\&\frac{\partial ^2 S}{\partial \gamma _0^2 } =\frac{\partial ^2 S_0}{\partial \gamma _0^2 }G_0-2\frac{\partial S}{ \partial \gamma _0}\frac{\partial S_0}{ \partial \gamma _0}G_1, \\&\frac{\partial ^2 S}{\partial \gamma _0 \partial \gamma _1} =\frac{\partial ^2 S_0}{\partial \gamma _0 \partial \gamma _1}G_0-2\frac{\partial S}{ \partial \gamma _0}\frac{\partial S_0}{ \partial \gamma _1}G_1 , ~~~~ \frac{\partial ^2 S}{\partial \gamma _1^2}=\frac{\partial ^2 S_0}{\partial \gamma _1 ^2}G_0-2\frac{\partial S}{ \partial \gamma _1}\frac{\partial S_0}{ \partial \gamma _1}G_1, \\&\frac{\partial ^2 S}{\partial \gamma _0 \partial \alpha _l}=\frac{\partial S }{\partial \gamma _0}x_{il}G_2,~~~ \frac{\partial ^2 S}{\partial \gamma _1 \partial \alpha _l}= \frac{\partial S}{ \partial \gamma _1} x_{il} G_2 ,~~~ \frac{\partial ^2 S}{\partial \alpha _l\partial \alpha _{l'}}=\frac{\partial S}{\partial \alpha _l}x_{il'}G_2; \end{aligned}$$

the derivatives of \(\log S\) are given by

$$\begin{aligned}&\frac{\partial \mathrm {log}S}{ \partial \gamma _0}=\frac{\partial \mathrm {log}S_0}{ \partial \gamma _0} \frac{1}{G } , ~~~~ \frac{\partial \mathrm {log} S}{ \partial \gamma _1} =\frac{\partial \mathrm {log}S_0}{ \partial \gamma _1} \frac{1}{G} , \\&\frac{\partial ^2 \mathrm {log}S}{\partial \gamma _0 ^2} =\left( \frac{\partial ^2 \mathrm {log}S_0}{\partial \gamma _0^2} -\frac{\partial \mathrm {log}S_0}{\partial \gamma _0} \frac{\partial S_0}{\partial \gamma _0} G_1\right) \frac{1}{G} ,\\&\frac{\partial ^2 \mathrm {log}S}{\partial \gamma _0 \partial \gamma _1} =\left( \frac{\partial ^2 \mathrm {log}S_0}{\partial \gamma _0\partial \gamma _1} -\frac{\partial \mathrm {log}S_0}{\partial \gamma _0} \frac{\partial S_0}{\partial \gamma _1} G_1\right) \frac{1}{G} , \\&\frac{\partial ^2 \mathrm {log}S}{\partial \gamma _1^2 }= \left( \frac{\partial ^2 \mathrm {log}S_0}{\partial \gamma _1^2} -\frac{\partial \mathrm {log}S_0}{\partial \gamma _1} \frac{\partial S_0}{\partial \gamma _1} G_1\right) \frac{1}{G} ,\\&\frac{\partial ^2 \mathrm {log}S}{\partial \gamma _0 \partial \alpha _{il}}=-\frac{\partial S}{\partial \gamma _0} x_{il} ,~~~ \frac{\partial ^2 \mathrm {log}S}{\partial \gamma _1 \partial \alpha _{il}}=- \frac{\partial S}{ \partial \gamma _1} x_{il} , \\&\frac{\partial \mathrm {log}S}{\partial \alpha _{il}}= \frac{x_{il}F_0}{G} ,~~~ \frac{\partial ^2 \mathrm {log}S}{\partial \alpha _{il}\partial \alpha _{il'}}=- \frac{\partial S}{\partial \alpha _{il}} x_{il'}, \end{aligned}$$

where \( G=1+S_0(y_ie^{\pmb {x}_i'\gamma _1}-1), \)

\(G_0=\frac{f}{f_0};~~G_1=\frac{y_ie^{\pmb {\beta }'\pmb {z}_i}-1}{G}, ~~ G_2= 2\frac{F_0}{G}-1, ,~~ G_3=(\gamma _1 t_i)^{1/\gamma _0} S_0G_1+1;\)

and finally, the derivatives of \(\log f_0\) are given by

$$\begin{aligned} \frac{\partial \mathrm {log}f}{\partial \gamma _0}&= \frac{\partial \mathrm {log} f_0}{\partial \gamma _0} -2 \frac{\partial S_0}{\partial \gamma _0} G_1, ~~ \frac{\partial \mathrm {log} f}{\partial \gamma _1} = \frac{\partial \mathrm {log} f_0}{\partial \gamma _1} -2 \frac{\partial S_0}{\partial \gamma _1} G_1, \\ \frac{\partial ^2\mathrm {log} f}{\partial \gamma _0^2}&= \frac{\partial ^2\mathrm {log}f_0}{\partial \gamma _0^2}-2 \frac{\partial ^2 S_0}{\partial \gamma _0^2} G_1+2 \left( \frac{\partial S_0}{\partial \gamma _0} G_1\right) ^2, \\ \frac{\partial ^2 \mathrm {log} f}{\partial \gamma _0 \partial \gamma _1}&= \frac{\partial ^2 \mathrm {log}f_0}{\partial \gamma _0 \partial \gamma _1}-2 \frac{\partial ^2 S_0}{\partial \gamma _0\partial \gamma _1} G_1+2 \frac{\partial S_0}{\partial \gamma _0}\frac{\partial S_0}{\partial \gamma _1}( G_1)^2, \\ \frac{\partial ^2 \mathrm {log} f}{\partial \gamma _0 \partial \alpha _l}&=-2 \frac{\partial S_0}{\partial \gamma _0} \frac{ x_{il}y_ie^{\pmb {\alpha }'\pmb {x}_i} }{G^2}, \\ \frac{\partial ^2 \mathrm {log} f}{\partial \gamma _1^2}&= \frac{\partial ^2 \mathrm {log}f_0}{\partial \gamma _1^2}-2 \frac{\partial ^2 S_0}{\partial \gamma _1^2} G_1+2 \left( \frac{\partial S_0}{\partial \gamma _1} G_1\right) ^2, \\ \frac{\partial ^2 \mathrm {log} f}{\partial \gamma _1 \partial \alpha _l}&=-2 \frac{\partial S_0}{\partial \gamma _1} \frac{ x_{il}y_ie^{\pmb {\alpha }'\pmb {x}_i}}{G^2} ,\\ \frac{\partial \mathrm {log}f}{\partial \alpha _l}&= x_{il}\left( 2\frac{ 1-S_0}{G}-1\right) , ~~~ \frac{\partial ^2 \mathrm {log}f}{\partial \alpha _l \partial \alpha _{l'}}=- \frac{2x_{il}x_{il'} S_0F_0y_ie^{\pmb {\alpha }'\pmb {x}_i}}{G^2 }. \end{aligned}$$

Expressions for Log-Logistic Baseline

In this case, we have the following expressions:

$$\begin{aligned} S_0=\frac{\gamma _0^{\gamma _1}}{t_i^{\gamma _1}+\gamma _0^{\gamma _1}} ,&~~ S=\frac{\gamma _0^{\gamma _1}y_ie^{\pmb {\alpha }'\pmb {x}}}{y_i \gamma _0^{\gamma _1}e^{\pmb {\alpha }'\pmb {x}}+t_i^{\gamma _1}}, ~~ \nonumber \\ f_0= \frac{\gamma _0^{\gamma _1}\gamma _1t_i^{\gamma _1-1}}{(t_i^{\gamma _1}+\gamma _0^{\gamma _1})^2} ,&~~ f= \frac{\gamma _0^{\gamma _1}\gamma _1t_i^{\gamma _1-1}e^{\pmb {\alpha }'\pmb {x}_i}}{(t_i^{\gamma _1}+\gamma _0^{\gamma _1}e^{\pmb {\alpha }'\pmb {x}_i})^2}. \end{aligned}$$
(49)

The derivatives of \(S(t_i,\pmb {\gamma })\) are given by,

$$\begin{aligned} \frac{\partial S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}&=F_0S_0\frac{\gamma _1}{\gamma _0} ,~~~ \frac{\partial S_0(t_i;\pmb {\gamma })}{\partial \gamma _1}= F_0 S_0\mathrm {log}\frac{\gamma _0}{t_i} ,\\ \frac{\partial ^2 S_0(t_i;\pmb {\gamma })}{\partial \gamma _0 ^2}&=-\frac{\partial S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}\frac{1+\gamma _1S_0(t_i;\pmb {\gamma })}{\gamma _0}+F_0S_0\frac{\gamma _1}{\gamma _0}F_0\frac{\gamma _1}{\gamma _0} ,\\ \frac{\partial ^2 S_0(t_i;\pmb {\gamma })}{\partial \gamma _0 \partial \gamma _1}&=\frac{\partial S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}\left( \frac{1}{\gamma _1}+S_0(t_i;\pmb {\gamma })\mathrm {log}\frac{t_i}{\gamma _0}\right) +F_0S_0\frac{\gamma _1}{\gamma _0} S_0\mathrm {log}\frac{\gamma _0}{t_i} ,\\ \frac{\partial ^2 S_0(t_i;\pmb {\gamma })}{\partial \gamma _1 ^2}&=\frac{\partial S_0(t_i;\pmb {\gamma })}{\partial \gamma _1 }S_0(t_i;\pmb {\gamma })\mathrm {log}\frac{t_i}{\gamma _0}+F_0 \mathrm {log}\frac{\gamma _0}{t_i}F_0 S_0\mathrm {log}\frac{\gamma _0}{t_i}; \end{aligned}$$

the derivatives of \(\mathrm {log} S(t_i,\pmb {\gamma })\) are given by

$$\begin{aligned} \frac{\partial \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}&=\frac{t_i^{\gamma _1}(\gamma _1/\gamma _0)}{\gamma _0^{\gamma _1}+t_i^{\gamma _1}} ,~~~ \frac{\partial \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _1}= \frac{t_i^{\gamma _1}\mathrm {log}(\gamma _0/t_i)}{\gamma _0^{\gamma _1}+t_i^{\gamma _1}} ,\\ \frac{\partial ^2\mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _0^2}&=-\frac{\partial \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}\frac{1+\gamma _1S_0(t_i;\pmb {\gamma })}{\gamma _0} ,~~~ \\ \frac{\partial \mathrm {log}S_0^2(t_i;\pmb {\gamma })}{\partial \gamma _0 \partial \gamma _1}&=\frac{\partial \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _0}\left( \frac{1}{\gamma _1}+S_0(t_i;\pmb {\gamma })\mathrm {log}\frac{t_i}{\gamma _0}\right) ,\\ \frac{\partial ^2 \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _1 ^2}&=\frac{\partial \mathrm {log}S_0(t_i;\pmb {\gamma })}{\partial \gamma _1 }S_0(t_i;\pmb {\gamma })\mathrm {log}\frac{t_i}{\gamma _0}; \end{aligned}$$

the derivatives of \(\mathrm {log} f(t_i,\pmb {\gamma })\) are given by

$$\begin{aligned}&\frac{\partial \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _0}=\frac{\gamma _1}{\gamma _0}\frac{t_i^{\gamma _1}-\gamma _0^{\gamma _1}}{\gamma _0^{\gamma _1}+t_i^{\gamma _1}} ,~~ \frac{\partial \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _1}=\frac{\gamma _0^{\gamma _1}-t_i^{\gamma _1}}{\gamma _0^{\gamma _1}+t_i^{\gamma _1}}\mathrm {log}\frac{t_i}{\gamma _0}+\frac{1}{\gamma _1} ,\\&\frac{\partial ^2\mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _0^2}=-\frac{\partial \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _0}\frac{1}{\gamma _0}-\frac{2\gamma _0^{\gamma _1}t_i^{\gamma _1}}{(\gamma _0^{\gamma _1}+t_i^{\gamma _1})^2}\frac{\gamma _1^2}{\gamma _0^2} ,\\&\frac{\partial ^2 \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _0\partial \gamma _1}=\frac{\partial \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _0}\frac{1}{\gamma _1}+\frac{2\gamma _0^{\gamma _1}t_i^{\gamma _1}}{(\gamma _0^{\gamma _1}+t_i^{\gamma _1})^2}\frac{\gamma _1}{\gamma _0}\mathrm {log}\frac{t_i}{\gamma _0} ,\\&\frac{\partial ^2 \mathrm {log} f_0(t_i,\pmb {\gamma })}{\partial \gamma _1^2}=-\frac{2x_{il}\gamma _0^{\gamma _1}t_i^{\gamma _1}}{(\gamma _0^{\gamma _1}+t_i^{\gamma _1})^2}\left( \mathrm {log}\frac{t_i}{\gamma _0}\right) ^2-\frac{1}{\gamma _1^2}; \end{aligned}$$

the derivatives of \( S(t_i|y_i)\) are given by

$$\begin{aligned} \frac{\partial S(t_i)}{\partial \gamma _0}&=F(t_i)S(t_i)\frac{\gamma _1}{\gamma _0} ,~~~ \frac{\partial S(t_i)}{\partial \gamma _1}=-S(t_i) F(t_i)\mathrm {log}\frac{t_i}{\gamma _0} ,~~~ \frac{\partial S(t_i)}{\partial \alpha _{l}}=x_{il}F(t_i)S(t_i) ,\\ \frac{\partial ^2 S(t_i)}{\partial \gamma _0^2}&=\frac{\partial S(t_i)}{\partial \gamma _0} \frac{\gamma _1(F(t_i)-S(t_i))-1}{\gamma _0} ,~~~\\ \frac{\partial S^2(t_i)}{\partial \gamma _0 \partial \gamma _1}&=\frac{\partial S(t_i)}{\partial \gamma _0}\left( \frac{1}{\gamma _1}-(F(t_i)-S(t_i))\mathrm {log}\frac{t_i}{\gamma _0}\right) ,\\ \frac{\partial ^2 S(t_i)}{\partial \gamma _0 \partial \alpha _{h}}&=\frac{\partial S(t_i)}{\partial \gamma _0}(F(t_i)-S(t_i))x_{il} ,~~~~~~~ \frac{\partial ^2 S(t_i)}{\partial \gamma _1^2}=\frac{\partial S(t_i)}{\partial \gamma _1 }(S(t_i)-F(t_i))\mathrm {log}\frac{t_i}{\gamma _0} , \\ \frac{\partial ^2 S(t_i)}{\partial \gamma _1 \partial \alpha _{h}}&=\frac{\partial S(t_i)}{\partial \gamma _1}(F(t_i)-S(t_i))x_{il} ,~~~~~~~ \frac{\partial ^2 S(t_i)}{\partial \alpha _{l} \partial \alpha _{l'}}= \frac{\partial S(t_i)}{\partial \alpha _{l} }(F(t_i)-S(t_i))x_{il'}; \end{aligned}$$

the derivatives of \(\mathrm {log} S(t_i|y_i)\) are given by

$$\begin{aligned} \frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _0}&=F(t_i)\frac{\gamma _1}{\gamma _0} ,~~~ \frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _1}= F(t_i)\mathrm {log}\frac{\gamma _0}{t_i} ,~~~ \frac{\partial \mathrm {log}S(t_i)}{\partial \alpha _{l}}=x_{il}F(t_i) ,\\ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \gamma _0^2}&=-\frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _0}\frac{1+\gamma _1S(t_i)}{\gamma _0} ,~~~ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \gamma _0 \partial \gamma _1}=\frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _0}\left( \frac{1}{\gamma _1}+S(t_i)\mathrm {log}\frac{t_i}{\gamma _0}\right) ,\\ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \gamma _0 \partial \alpha _{h}}&=-\frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _0}S(t_i)x_{il} ,~~~~~~~ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \gamma _1^2}=\frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _1 }S(t_i)\mathrm {log}\frac{t_i}{\gamma _0} , \\ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \gamma _1 \partial \alpha _{l}}&=-\frac{\partial \mathrm {log}S(t_i)}{\partial \gamma _1}S(t_i)x_{il} ,~~~~~~~ \frac{\partial ^2 \mathrm {log}S(t_i)}{\partial \alpha _{l} \partial \alpha _{l'}}=-x_{il'} \frac{\partial S(t_i)}{\partial \alpha _{l} }; \end{aligned}$$

and finally, the derivatives of \(\mathrm {log} f(t_i|y_i)\) are given by

$$\begin{aligned}&\frac{\partial \mathrm {log} f(t_i)}{\partial \gamma _0}=\frac{\gamma _1}{\gamma _0}V_i ,~~ \frac{\partial \mathrm {log} f(t_i)}{\partial \gamma _1}=\frac{1}{\gamma _1}-V_i\mathrm {log}\frac{t_i}{\gamma _0} ,~~~ \frac{\partial \mathrm {log} f(t_i)}{\partial \alpha _{l}}=x_{il}V_i ,\\&\frac{\partial ^2 \mathrm {log} f(t_i)}{\partial \gamma _0^2}=-\frac{\partial \mathrm {log} f(t_i)}{\partial \gamma _0}\frac{1}{\gamma _0}-W_i\frac{\gamma _1^2}{\gamma _0^2} ,~~~ \frac{\partial ^2 \mathrm {log} f(t_i)}{\partial \gamma _0\partial \gamma _1}=\frac{\partial \mathrm {log} f(t_i)}{\partial \gamma _0}\frac{1}{\gamma _1}+W_i\frac{\gamma _1}{\gamma _0}\mathrm {log}\frac{t_i}{\gamma _0} ,\\&\frac{\partial ^2 \mathrm {log} f(t_i)}{\partial \gamma _0\partial \alpha _{l}}=-W_i\frac{\gamma _1}{\gamma _0}x_{il} ,~~~ \frac{\partial ^2 \mathrm {log} f(t_i)}{\partial \gamma _1^2}=-W_i\left( \mathrm {log}\frac{t_i}{\gamma _0}\right) ^2-\frac{1}{\gamma _1^2} ,\\&\frac{\partial ^2 \mathrm {log} f(t_i)}{\partial \gamma _1\partial \alpha _{l}}=W_i\mathrm {log}\frac{t_i}{\gamma _0}x_{il} ,~~~ \frac{\partial \mathrm {log} f^2(t_i)}{\partial \alpha _{l}\partial \alpha _{l'}}=-x_{il}x_{il'}W_i, \end{aligned}$$

where \(V_i=F(t_i)-S(t_i)\),and \(W_i=\frac{2\gamma _0^{\gamma _1}y_ie^{\pmb {\alpha }'\pmb {x}}t_i^{\gamma _1}}{(\gamma _0^{\gamma _1}y_ie^{\pmb {\alpha }'\pmb {x}}+t_i^{\gamma _1})^2}\).

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Check for updates. Verify currency and authenticity via CrossMark

Cite this chapter

Balakrishnan, N., Feng, T., So, HY. (2023). Proportional Odds COM-Poisson Cure Rate Model with Gamma Frailty and Associated Inference and Application. In: Balakrishnan, N., Gil, M.Á., Martín, N., Morales, D., Pardo, M.d.C. (eds) Trends in Mathematical, Information and Data Sciences. Studies in Systems, Decision and Control, vol 445. Springer, Cham. https://doi.org/10.1007/978-3-031-04137-2_23

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-04137-2_23

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-04136-5

  • Online ISBN: 978-3-031-04137-2

  • eBook Packages: EngineeringEngineering (R0)

Publish with us

Policies and ethics