A plug-in rule for bandwidth selection in circular density estimation

TL;DR

This paper introduces a new plug-in rule for selecting bandwidths in kernel circular density estimation, demonstrating its effectiveness through simulations and real data applications.

Contribution

The paper proposes a novel plug-in bandwidth selection method specifically for circular density estimation, improving upon existing methods.

Findings

The new rule performs well across various circular distributions.

It outperforms some existing bandwidth selectors in simulations.

The method is successfully applied to real datasets.

Abstract

A new plug-in rule procedure for bandwidth selection in kernel circular density estimation is introduced. The performance of this proposal is checked throughout a simulation study considering a variety of circular distributions exhibiting multimodality, peakedness and/or skewness. The plug-in rule behaviour is also compared with other existing bandwidth selectors. The method is illustrated with two classical datasets of cross-beds layers and animal orientation.