This paper presents Modified Gauss-Newton iteration algorithm for the nonlinear regression models for Failure Time Data set. The convergence of the iteration is proved carefully. Simulation illustrated that our method is available. Our results may be regarded as an extension of Wei (1998) for exponential nonlinear regression models without failure time data.
AitkinM.A note on the regression analysis of censored data, Technometrics, 23, 161–163 (1981).
2.
CaiT.ChengS. C. and WeiL. J.Semiparametric Mixed-effects Models for Cluster Failure Time Data, Journal of American Statistical Association, 97, 514–522 (2002).
3.
CookR. D. and TsaiC. L.Residuals in nonlinear regression. Biometrika, 72, 23–29 (1985).
4.
CoxD. R.Regression Models and Life Tables (with discussion). Journal of the Royal Statistical Society, Ser. B, 34, 187–220 (1972).
5.
HamiltonB. C.Confidence regions for parameter subsets in nonlinear regression. Biometrika, 73, 57–64 (1986).
6.
HougaardP.A class of Multivariate Failure Time Distributions, Biometrika, 73, 671–678 (1986).
7.
KleinJ. P.Semiparametric Estimation of Random effects using Cox model based on EM algorithm, Biometrics, 48, 795–806 (1992).
8.
ZhangD.LinX.RazJ. and SowersM.Semiparametric Stochastic Mixed Models for longitudinal. Journal of American Statistical Association, 93, 710–719 (1998).
9.
WeiB. C., Exponential Family Nonlinear Models.Singapore: Springer-Verlag Singapore (1998).
10.
YuQ. Q. and WongG. Y. C.Asymptotic Properties of a modified Semi-parametric MLE in linear Regression with Right-censored Data, Acta Mathematics Sinica, 18, 405–416 (2002).
11.
Zongxu-Ping. Confidence on nonlinear models with random effects, Applied Mathematics, Vol. 13, No. 4., 100–105 (2000).
12.
Xu-PingZong. Estimation in linear models with random effects and errors-invariables, Ann. Inst. Stat. Math., Vol. 54, NO. 3, pp595–606 (2002).
