HIGHLIGHTS
- who: Aki Vehtari and colleagues from the Center for Computational Mathematics, Flatiron Institute, New York ** Department of Computer Science, Aalto University, Finland have published the research: Rank-Normalization, Folding, and Localization: An Improved for Assessing Convergence of MCMC (with Discussion), in the Journal: (JOURNAL)
- what: The authors show that the convergence diagnostic will fail to correctly Gelman and Rubin has serious flaws. The authors propose alternative rank-based diagnostic that fixes these problems. This paper proposes improvements that overcome these problems. As the convergence of the Markov chain needs not be uniform across the parameter . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.