A note on kernel density estimation for undirected dyadic data
证明在无向二元网络中，结果密度的估计量以√N速率收敛到正态分布，该结论基于Frees的U统计量经典结果，并简化了Graham等人的推导。
.In this note, I show that the √𝑁 convergence to the normal distribution holds for the density of outcomes generated from a dyadic network using the seminal result in the U-statistic literature obtained by Frees. In particular, our derivations imply that the main result for the non degenerate case in Graham, Niu, and Powell follows from arguments in Frees.