博士生讨论班2024[06]
时间：2024-10-24
阅读量：348次
- 演讲人： 陈书源
- 时间：2024年10日29日14:00
- 地点：浙江大学紫金港校区行政楼1417报告厅
报告主题：Doubly robust estimation in partial linear Cox model
摘要：The partial linear Cox model is a powerful extension to the Cox proportional hazards model for survival analysis. It facilitates the statistical inference on important covariates in the deep learning framework. However, as a price, the flexibility and generality brought by estimating nuisance function usually accompany biases in estimating the parameters of interest. A doubly robust estimator is proposed in this paper to reduce the biases caused by fitting nuisance functions with some flexible nonparametric functions. It is obtained by solving a doubly robust estimating equation, designed particularly for the partial linear Cox model, based on the counting process theory. We present the statistical properties of this doubly robust estimator, including consistency and asymptotic normality.