File "zero"

 This file provides a bayes classifier with routines to determine binary decision diagrams and their failure score. The bayes classifier can find solutions for multi-output problems. A solution is obtained by performing a naïve statistic for each output and then using a majority rule. Ties or missing counts are resolved by taking the decision 0.

The following zero predicates are provided:

zero_bayes(N, P-Q, G, C):
The predicate succeeds in C with the binary decision diagrams, that map P to Q from G obtained by the majority rule. The parameter N gives the ensemble size.
zero_failure(P-Q, G, C, S):
The predicate succeeds in S with the failure counts, comparing the binary decision  diagrams C to the map P to Q from G.


Kommentare