单调回归模型及其在退化数据中的应用

The Estimation of the Monotone Regression Function and Its Application to the Analysis of the Degradation Data

  • 摘要: 形如f=C_1+C_2 \int_0^t\left\exp \int_0^u \omega(s) \mathrmd s\right \mathrmd u(其中:C1,C2是常数,w(s)是L2可积)的函数是一个广泛的单调函数类.特别地,假定w(s)是有固定结点的样条函数.我们用惩罚最小二乘的方法估计其中的未知参数,得到了参数估计的强相合性及渐近正态性,并将该模型用到退化实例中去,得到很好的效果.

     

    Abstract: The equation such asf=C_1+C_2 \int_0^t\left\exp \int_0^u \omega(s) \mathrmd s\right \mathrmd u, where C1,C2 are arbitrary constants and w(s) is Lebesgue square integrable, is the famility of monotone function. In particular the function w(s) is supposed to be a spline function in this paper. The parameters are estimated by the penalized least square method and the asymptotic of the estimator is obtained. Application is discussed to the analysis of the degradation data and the satisfying result is made.

     

/

返回文章
返回