This is one of the most employed kernels, whose structure is as follows:
The graphical representation is shown in the following screenshot:
Gaussian kernel
Given its regularity, a Gaussian kernel is a common choice for many density estimation tasks. However, as the method doesn't allow the mixing of different kernels, the choice must consider all the properties. From statistics, we know that a Gaussian distribution can be considered as an average reference for the kurtosis (which is proportional to the peak and to the heaviness ...
