Procedure to automatically eliminate hypotheses based on integrated trial results aaa

Incorrect hypotheses are not explicitly removed at any stage of the
discovery cycle, rather the collection of more evidence will result in
some hypotheses no longer being consistent with all of the available
evidence. The aim of experiment selection is to choose an experiment
that will refute a large number of candidate hypothesis regardless of
the actual outcome of the experiment. Hypotheses generation by
C-Progol only generates those hypotheses consistent with the
background theory and the set of training examples.