Previous lecture shows how to generate random variable from beta density, and more generally from ANY density.
This lecture shows how GIVEN A GENERATED data set, HOW can we found out/estimate the DENSITY from which the data was generated
Basically, we use Kernel Density Estimators. To get good estimates by this method, it is essential to vary the window size. However this technique is not taught, and not implemented in computer packages.
Representing Data via Kernel Density Estimates - ADVANCED VERSION - 1 hr 25 min - Covers many additional topic (which are not relevant for BAYESIAN course)
Lecture Notes and Supplementary Materials - Kernel Density Estimation -- with variable window size algorithm implemented in EXCEL
Kernel Density Estimators from Data (Basic Introduction) - 1 hr 15 min video lecture