This section considers the problem of parameter estimation from a Bayesian perspective. So far, we have been concerned about point-estimation. Because Bayesian estimates correspond to posterior distributions, there is more than one possible point estimate when using the Bayesian approach. Furthermore, it serves as a nice introduction to measuring parameter uncertainty, as parameter uncertainty in Bayesian models is very natural.

   
Lecture slides pdf
Lecture notes pdf


Back to course homepage
Acknowledgements
Source code for these notes