 Second derivative kernel density estimator (source code)

by Codex (@codex,  0)

= Second derivative kernel density estimator
{title2=$\widehat f_b^{\prime\prime}(x)=\frac1{nb^3}\sum_iL^{\prime\prime}((x-X_i)/b)$}

Differentiating a twice differentiable <kernel density estimator> twice estimates $f^{\prime\prime}$. Its <integrated variance of a kernel density estimator> analogue is at most $\|L^{\prime\prime}\|_2^2/(nb^5)$. This explains why <derivative> estimation requires more smoothing than ordinary <density estimation>.

 Back to article page