相依样本条件t-分位数核估计的强相合性

The strong consistency of conditional t-quantiles kernel estimate for dependent sample

对φ-混合样本下条件t-分位数θx(t)的核估计θx,n(t),取消核函数的支撑为紧集的限制,同时放宽对混合系数的要求,研究了其强收敛性,该结果拓宽了核估计θx,n(t)的适用范围.

For conditional t-quantiles kernel estimate of φ-mixing sample, in this paper, its strong consistency is studied with canceling the restriction the compactness of support set of kernel function, and with widening the limit to the mixing coefficient. This result extends the applied scope of kernel estimate.