Here are files concerning the next-generation of weighted-likeihood bootstrap computations.
We seek conditions and numerical examples that demonstrate how repeated optimization of a suitably randomized log-(joint)-posterior gives a sample of optimizers that approximate a bona fide sample from the corresponding posterior distribution.
See our latest paper at the Electronic Journal of Statistics.