The estimate \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$f(\hat {r}_{\text {\textit {j,k}}})$\end{document}f(r^j,k) is still a biased estimator of δjk. Thus, we propose a second step to remove the bias. Using the Taylor series expansion, we can estimate the bias as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document} $$\frac{c_{1}}{n}\left(1-\hat{r}_{j,k}^{2} \right)^{2}. $$ \end{document}c1n1−r^j,k22.
