带有测量误差变系数模型的模型选择

Model Selection for Varying-Coefficient Models with Measurement Errors

本文提出了一种惩罚的偏差校正方法以确定带有测量误差的变系数模型的真实结构,并利用局部二次逼近算法给出估计值的计算。通过数值糢拟验证了本文所提出方法的有限样本表现.最后,将我们的方法应用到艾滋病数据的分析中.

This paper proposes a bias-corrected penalized method to determine the underlying true structure of varying-coefficient models with measurement errors.Using the local quadratic approximations,we give the algorithm to compute the estimates.Simulation studies are conducted to assess the finite sample performance of the proposed method.A real data application shows the performance of our method in the end.