平稳过程的条件概率密度非参数估计的L1－模强相合性

Almost Sure L1-Norm Convergence of Nonparametric Estimation of Conditional Probability Densities of Stationary processes

设｛Ｘｊ｝∞ｊ＝－∞是一个实值平稳随机过程，本文考察了联合密度和给定过去状态的条件密度的递推估计，得到这些估计在过程｛Ｘｊ｝的各种混合条件下是Ｌ１－模强相合的。

Let be a real-valued statonary process. Recursive kernel estimators of the joint probability density functions, of conditional probability densities, given past behavior, are considered. Their almost sure L1-norm convergences are established for processes {Xj }satisfying various mixing conditions.