1: \begin{abstract}

2: We consider the problem of estimating the $ s $-th derivative of a density function $ f $ by the \textit{tilted} Kernel estimator introduced in \cite{Hall_Doosti}. Then we further show this estimator achieves the same convergence rate, in probability, the wavelet estimators achieved as shown in \cite{Hall_Patil_1995}. That is, the convergence rate of $ O_p \left( n^{-(r-s)/(2r+1)} \right) $.

3: \end{abstract}

4: