Frequentist MCMC

<p>Some definitions are introduced and exemplified that may help to relate Bayesian statistics to frequentist statistics. The idea is interesting. More work is required.</p> <p> </p> <p>Practical implications would be:</p> <p>- Opens up the possibility to use MCMC algorithms sampling parameters given data, e.g., Stan or WinBUGS, for frequentist hypothesis testing.</p> <p> </p> <p>Conceptual implications would be:</p> <p>- Formally relate results from Bayesian statistics to those from frequentist statistics.</p> <p>- Define approaches and situations, in which frequentist and Bayesian approaches give identical results, or to explain differences obtained with different approaches.</p> <p> </p> <p>References: </p> <p>- Kass RE, Wasserman L (1996) The Selection of Prior Distributions by Formal Rules. JASA, 91 (435), 1343-1370</p> <p>- Berger, James. "The case for objective Bayesian analysis." Bayesian Analysis 1.3 (2006): 385-402.</p> <p>- Bernardo, José M. "Intrinsic credible regions: An objective Bayesian approach to interval estimation." Test 14.2 (2005): 317-384.</p> <p>- Efron, Bradley. "A 250-year argument: Belief, behavior, and the bootstrap." Bulletin of the American Mathematical Society 50.1 (2013): 129-146.</p> <p>- Berger, James O, Brunero Liseo, and Robert L Wolpert. "Integrated likelihood methods for eliminating nuisance parameters." Statistical Science 14.1 (1999): 1-28.</p> <p>- Severini, Thomas A. "Integrated likelihood functions for non-Bayesian inference." Biometrika 94.3 (2007): 529-542.</p> <p>- Berger, James O, José M Bernardo, and Dongchu Sun. "The formal definition of reference priors." The Annals of Statistics (2009): 905-938.</p> <p>- Bjørnstad, Jan F. "On the generalization of the likelihood function and the likelihood principle." Journal of the American Statistical Association 91.434 (1996): 791-806.