README
¶
decisiontrees
Gradient Boosting, Random Forests, etc. implemented in Go
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func LearningCurve ¶
func LearningCurve(f *pb.Forest, e Examples) *pb.TrainingResults
LearningCurve computes the progressive learning curve after each epoch on the given examples
Types ¶
type EvaluatorFunc ¶
EvaluatorFunc implements the Evaluator interface
func (EvaluatorFunc) Evaluate ¶
func (f EvaluatorFunc) Evaluate(features []float64) float64
Evaluate is the implementation of Evaluator interface
type FeatureSelector ¶
type FeatureSelector interface {
// contains filtered or unexported methods
}
FeatureSelector allows algorithms to configure which features to use for a given round of splitting
type ForestGenerator ¶
ForestGenerator is implemented by various algorithms that generate an ensemble of decision trees from the given training dataset.
func NewForestGenerator ¶
func NewForestGenerator(forestConfig *pb.ForestConfig) (ForestGenerator, error)
NewForestGenerator returns a ForeestGenerator from the given ForestConfig.
type LossFunction ¶
type LossFunction interface {
UpdateWeightedLabels(e Examples)
GetPrior(e Examples) float64
GetLeafWeight(e Examples) float64
GetSampleImportance(ex *pb.Example) float64
}
LossFunction is an arbitrary loss function used in computing decision trees
func NewLossFunction ¶
func NewLossFunction(l *pb.LossFunctionConfig, evaluator Evaluator) LossFunction
NewLossFunction returns an implementation of `LossFunction` given the LossFunctionConfig
Source Files
Directories