This paper deals with a nonparametric Nadaraya-Watson estimator $\widehat b$ of the drift function computed from independent continuous observations of a diffusion process. Risk bounds on $\widehat b$ and its discrete-time approximation are established. The paper also deals with extensions of the PCO and leave-one-out cross validation bandwidth selection methods for $\widehat b$. Finally, some numerical experiments are provided.
翻译:本文涉及从对扩散过程的独立连续观测中计算出来的漂移函数的非参数Nadaraya-Watson估计值$\全方位 b$; 设定了对美元全方位b$及其离散时间近似值的风险约束; 本文还涉及PCO的扩展和以$全方位b$的放出一放的跨频带宽选择方法; 最后,提供了一些数字实验。