bart tree parameters More...

#include <BART.h>

Public Member Functions
	bart_params ()
	a simple default ctor

void	set_classification (int num_trees)
	an initializer for classification problems

void	set_regression (int num_trees, float sample_var_y_in_data)
	an initializer for regression problems.

Data Fields
int	min_obs_in_node
	minimal allowed observations in node

float	alpha
	prior for tree structure: alpha * (1 + depth(node)) ^ -beta

float	beta
	prior for tree structure: alpha * (1 + depth(node)) ^ -beta

bart_data_prior_type	data_prior_type

float	k
	the range for bandwidth interval

float	nu
	the node-data dict params for sigma_i: sigma_i ~ IG(nu, mean_sigma*lambda/2)

float	lambda
	the node-data dict params for sigma_i: sigma_i ~ IG(mean_sigma/2, mean_sigma*lambda/2)

Detailed Description

bart tree parameters

Member Function Documentation

void bart_params::set_regression	(	int	num_trees,
		float	sample_var_y_in_data
	)

inline

an initializer for regression problems.

Parameters

sample_var_y_in_data it's the variance in train labels

The documentation for this class was generated from the following file: