The first talk was very well prepared and delivered. It was accessible to most newbies, but I don't think M-H and Gibbs could be understood from the talk. There was some slight imprecision (acceptance-rejection was referred to "importance sampling": the former is a method to generate random variates, and the latter is a simulation scheme). Variational methods were new to most people, and the application was very very interesting. I would have preferred a talk less devoted to the basics, and more to i) advanced methods, and ii) their R implementation. I had to leave in the middle of the second talk. Engaging speaker. Very elementary, but easy to follow. Overall, a very well organized event. Thanks to Joshua and Mark!