Predictive Model is abstract class of predictor model which has parameters for GD or SGD uses it also has function to retrieve direct sub-gradients for the loss function of the model. More...

#include <PredictiveModel.h>

Inheritance diagram for PredictiveModel:

Public Member Functions
	PredictiveModel (string name)
	The name of the model.

virtual double	predict (const vector< float > &input) const =0

virtual subGradientFunction	getSubGradients ()
	Subgradient function to calc directly the gradient descent.

virtual void	predict (const vector< vector< float > > &inputs, vector< double > &preds) const
	virtual to allow more efficeint implemention

virtual void	print (const vector< string > &signalNames) const =0
	print model to stdout

virtual PredictiveModel *	clone () const =0
	copy model

Data Fields
vector< double >	model_params
	model parameters

string	model_name
	model name

Detailed Description

Predictive Model is abstract class of predictor model which has parameters for GD or SGD uses it also has function to retrieve direct sub-gradients for the loss function of the model.

Member Function Documentation

◆ clone()

virtual PredictiveModel * PredictiveModel::clone ( ) const

pure virtual

copy model

Implemented in MedLinearModel.

◆ getSubGradients()

subGradientFunction PredictiveModel::getSubGradients ( )

virtual

Subgradient function to calc directly the gradient descent.

Reimplemented in MedLinearModel.

◆ predict()

void PredictiveModel::predict	(	const vector< vector< float > > &	inputs,
		vector< double > &	preds
	)		const

virtual

virtual to allow more efficeint implemention

Reimplemented in MedLinearModel.

◆ print()

virtual void PredictiveModel::print ( const vector< string > & signalNames ) const

pure virtual

print model to stdout