Kernel density estimation (KDE) is a versatile nonparametric approach to infer continuous probability distributions from finite samples. By superimposing smooth kernel functions—most commonly Gaussian ...