Wald Lecture: On the Bernstein-von Mises theorem with infinite-dimensional parameters
If there are many independent, identically distributed\nobservations governed by a smooth, finite-dimensional statistical model, the\nBayes estimate and the maximum likelihood e...