摘要
线性模型Ｙｉ＝ｘ′ｉβ＋ｅｉ，ｉ＝１，２，…，其中ｅｉ｛｝∞ｉ＝１ｉｄ．，有未知密度ｆ（ｘ）．讨论了在对β作一般估计后，基于残差作出的误差密度核估计的相合性．在比文献［７，８］弱的条件下，证明了误差密度核估计的逐点弱相合，逐点强相合和一致强相合．
Linear model Y i=x′ iβ+e i, i=1,2,…, where e i ∞ i=1 iid. with unknow density f(x). In this paper, some convergences of kernel estimation are considered of error density based on the residuals of M estimation. Under weaker assumptions, the weak local convergence, the strong local convergence and the strong uniform convergence of kernel estimation of error density are given.
基金
国家自然科学基金
关键词
M-估计
残差
核估计
线性回归模型
误差密度
M estimator, residuals error, kernel estimation, convergenceA Method of Data Expansion Based on Wavelet Transform$$$$ SHAO Limin SHAO Xueguang (Department of Chemistry, USTC) Abstract A method of data expansion is established in this paper based on
