Dr. Peter Muller from the Department of Mathematics at the University of Texas. "A Bayes Rule for Subgroup Reporting."
Dr. Muller was the speaker at this week's seminar. And his entire talk was on the background to my project. In the presentation, Dr. Muller addresses various problems associated with design analysis and how using a Bayesian framework can help alleviate some of this problems. Subgroups can be picked up in a trial and then, using that preliminary information, determine to push the trial in a different way. In other words, using Bayesian statistics, researchers can change a trial in mid-operation.
As I am still trying to understand what my project will be, this talk was helpful, enabling me to sort through terminology and convoluted wording. I got a better grasp at my problem through this talk.
No comments:
Post a Comment