Adds a new trait, Voter, which is responsible of normalizing and combining predictions across multiple trees in an ensemble. This also introduces a new type parameter, P, which is used by both Voter and Error, for the combined prediction.
This also temporarily shelves BrierDecompositionError and brings back the earlier BrierScoreError, for now, until I get a chance to debug the former.
Adds a new trait,
Voter, which is responsible of normalizing and combining predictions across multiple trees in an ensemble. This also introduces a new type parameter,P, which is used by bothVoterandError, for the combined prediction.This also temporarily shelves
BrierDecompositionErrorand brings back the earlierBrierScoreError, for now, until I get a chance to debug the former.