Introduction to Kernel Density Estimation

So far, we've studied parametric distributions in this chapter, but in real life, all distributions are either approximations of parametric distributions or don't resemble any parametric distributions at all. In such cases, we use a technique called Kernel Density Estimation, or KDE, to estimate their probability distributions.

KDE is used to estimate the probability density function of distributions or random variables with given finite points of that distribution using something called a kernel. This will be more clear to you after you continue further in the chapter.

KDE Algorithm

Contrary to what it might seem like given the heavy name, KDE is a very simple two-step process:

1. Choosing a kernel

2. Placing the kernel on data points and taking the sum of kernels

A kernel is a non-negative symmetric function that is used to model distributions. For example, in KDE, a normal distribution function is the most commonly used kernel function. Kernel functions...