Predictive subset selection using imprecise probabilities

Following a recently introduced approach to statistical selection, new results are presented on subset selection. The inferences have a nonparametric predictive nature. The basic assumption is Hill's A(n). Assuming that values of some random quantities from k \geq 2 independent sources are observed, A(n) provides a predictive probabilistic inference for further unknown random quantities from each source. Selection is based on imprecise probabilities for the event that the next observation from at least one selected source will be greater than the next observation from every non-selected source. Results for selection of a single source are reviewed. New results are presented on subset selection, where selection is considered successful if the selected subset includes the best source.