File rule_evaluation.hpp

namespace boosting

template<typename StatisticVector>
class IRuleEvaluation

#include <rule_evaluation.hpp>

Defines an interface for all classes that allow to calculate the predictions of rules, as well as their overall quality, based on the gradients and Hessians that have been calculated according to a loss function.

Template Parameters:

StatisticVector – The type of the vector that provides access to the gradients and Hessians

Public Functions

inline virtual ~IRuleEvaluation()

virtual const IScoreVector &calculateScores(StatisticVector &statisticVector) = 0

Calculates the scores to be predicted by a rule, as well as its overall quality, based on the sums of gradients and Hessians that are covered by the rule.

Parameters:

statisticVector – A reference to an object of template type StatisticVector that stores the gradients and Hessians

Returns:

A reference to an object of type IScoreVector that stores the predicted scores of a rule, as well as its overall quality