drwnDecisionTree Class Reference

Implements a (binary-split) decision tree classifier of arbitrary depth. More...

Inheritance diagram for drwnDecisionTree:

Public Member Functions
	drwnDecisionTree ()
	default constructor
	drwnDecisionTree (unsigned n, unsigned k=2)
	construct a classifier object for n features and k classes
	drwnDecisionTree (const drwnDecisionTree &c)
	copy constructor
virtual const char *	type () const
	returns object type as a string (e.g., Foo::type() { return "Foo"; })
virtual drwnDecisionTree *	clone () const
	returns a copy of the class usually implemented as virtual Foo* clone() { return new Foo(*this); }
virtual void	initialize (unsigned n, unsigned k=2)
	initialize the classifier object for n features and k classes
virtual bool	save (drwnXMLNode &node) const
	write object to XML node (see also write)
virtual bool	load (drwnXMLNode &node)
	read object from XML node (see also read)
virtual double	train (const drwnClassifierDataset &dataset)
	train the parameters of the classifier from a drwnClassifierDataset object
virtual double	train (const vector< vector< double > > &features, const vector< int > &targets)
	train the parameters of the classifier from a set of features and corresponding labels
virtual double	train (const vector< vector< double > > &features, const vector< int > &targets, const vector< double > &weights)
	train the parameters of the classifier from a weighted set of features and corresponding labels
virtual void	getClassScores (const vector< double > &features, vector< double > &outputScores) const
	compute the unnormalized log-probability for a single feature vector
virtual int	getClassification (const vector< double > &features) const
	return the most likely class label for a single feature vector
Public Member Functions inherited from drwnClassifier
	drwnClassifier ()
	default constructor
	drwnClassifier (unsigned n, unsigned k=2)
	construct a classifer with n features and k classes
	drwnClassifier (const drwnClassifier &c)
	copy constructor
int	numFeatures () const
	returns the number of features expected by the classifier object
int	numClasses () const
	returns the number of classes predicted by the classifier object
virtual bool	valid () const
	returns true if the classifier is valid (has been initialized and trained)
virtual double	train (const char *filename)
	train the parameters of the classifier from data stored in filename
virtual void	getClassScores (const vector< vector< double > > &features, vector< vector< double > > &outputScores) const
	compute the unnormalized log-probability for a set of feature vectors
virtual void	getClassMarginals (const vector< double > &features, vector< double > &outputMarginals) const
	compute the class marginal probabilities for a single feature vector
virtual void	getClassMarginals (const vector< vector< double > > &features, vector< vector< double > > &outputMarginals) const
	compute the class marginal probabilities for a set of feature vectors
virtual void	getClassifications (const vector< vector< double > > &features, vector< int > &outputLabels) const
	compute the most likely class labels for a set of feature vector
Public Member Functions inherited from drwnWriteable
bool	write (const char *filename) const
	write object to file (calls save)
bool	read (const char *filename)
	read object from file (calls load)
void	dump () const
	print object's current state to standard output (for debugging)
Public Member Functions inherited from drwnProperties
unsigned	numProperties () const
bool	hasProperty (const string &name) const
bool	hasProperty (const char *name) const
unsigned	findProperty (const string &name) const
unsigned	findProperty (const char *name) const
void	setProperty (unsigned indx, bool value)
void	setProperty (unsigned indx, int value)
void	setProperty (unsigned indx, double value)
void	setProperty (unsigned indx, const string &value)
void	setProperty (unsigned indx, const char *value)
void	setProperty (unsigned indx, const Eigen::VectorXd &value)
void	setProperty (unsigned indx, const Eigen::MatrixXd &value)
void	setProperty (const char *name, bool value)
void	setProperty (const char *name, int value)
void	setProperty (const char *name, double value)
void	setProperty (const char *name, const string &value)
void	setProperty (const char *name, const char *value)
void	setProperty (const char *name, const Eigen::VectorXd &value)
void	setProperty (const char *name, const Eigen::MatrixXd &value)
string	getPropertyAsString (unsigned indx) const
drwnPropertyType	getPropertyType (unsigned indx) const
bool	isReadOnly (unsigned indx) const
const drwnPropertyInterface *	getProperty (unsigned indx) const
const drwnPropertyInterface *	getProperty (const char *name) const
bool	getBoolProperty (unsigned indx) const
int	getIntProperty (unsigned indx) const
double	getDoubleProperty (unsigned indx) const
const string &	getStringProperty (unsigned indx) const
const list< string > &	getListProperty (unsigned indx) const
int	getSelectionProperty (unsigned indx) const
const Eigen::VectorXd &	getVectorProperty (unsigned indx) const
const Eigen::MatrixXd &	getMatrixProperty (unsigned indx) const
const string &	getPropertyName (unsigned indx) const
vector< string >	getPropertyNames () const
void	readProperties (drwnXMLNode &xml, const char *tag="property")
void	writeProperties (drwnXMLNode &xml, const char *tag="property") const
void	printProperties (ostream &os) const

Static Public Attributes
static int	MAX_DEPTH = 1
	default maximum tree depth
static int	MAX_FEATURE_THRESHOLDS = 1000
	maximum number of thresholds to try during learning
static int	MIN_SAMPLES = 0
	minimum number of samples (after first split)
static double	LEAKAGE = 0.0
	probability that a training sample leaks to both splits
static drwnTreeSplitCriterion	SPLIT_CRITERION = DRWN_DT_SPLIT_ENTROPY
	tree split criteria during learning
static bool	CACHE_SORTED_INDEXES = true
	pre-cache indexes of sorted features

Protected Member Functions
const Eigen::VectorXd &	evaluate (const Eigen::VectorXd &x) const
void	learnDecisionTree (const vector< vector< double > > &x, const vector< int > &y, const vector< double > &w, const vector< vector< int > > &sortIndex, const drwnBitArray &sampleIndex)
Protected Member Functions inherited from drwnProperties
void	declareProperty (const string &name, drwnPropertyInterface *optif)
void	undeclareProperty (const string &name)
void	exposeProperties (drwnProperties *opts, const string &prefix=string(""), bool bSerializable=false)
virtual void	propertyChanged (const string &name)

Static Protected Member Functions
static void	computeSortedFeatureIndex (const vector< vector< double > > &x, const drwnBitArray &sampleIndex, int featureIndx, vector< int > &featureSortIndex)
static void	computeSortedFeatureIndex (const vector< vector< double > > &x, vector< vector< int > > &sortIndex)

Protected Attributes
int	_splitIndx
	variable index on which to split
double	_splitValue
	split value (go left if less than)
drwnDecisionTree *	_leftChild
	left child (or NULL)
drwnDecisionTree *	_rightChild
	right child (or NULL)
Eigen::VectorXd	_scores
	log-marginal for each class at this node
int	_predictedClass
	argmax of _scores
int	_maxDepth
	maximum depth of decision tree
Protected Attributes inherited from drwnClassifier
int	_nFeatures
	number of features
int	_nClasses
	number of classes
bool	_bValid
	true if the classifier has been trained or loaded

Friends
class	drwnDecisionTreeThread
class	drwnDecisionTreeConfig
class	drwnBoostedClassifier
class	drwnRandomForest

Detailed Description

Implements a (binary-split) decision tree classifier of arbitrary depth.

The following code snippet shows example learning a decision tree classifier on a training dataset and then testing it on a hold out evaluation dataset.

// load training dataset
drwnClassifierDataset dataset;
dataset.read("training_data.bin");

// train the classifier
const int nFeatures = dataset.numFeatures();
const int nClasses = dataset.maxTarget() + 1;
drwnDecisionTree model(nFeatures, nClasses);
model.train(dataset);

// load evaluation set
dataset.read("testing_data.bin", false);

// predict labels
vector<int> predictions;
model.getClassifications(dataset.features, predictions);

The decision classifier has a number of parameters for controlling it's operation during training. See drwnDecisionTree::MAX_DEPTH, drwnDecisionTree::MAX_FEATURE_THRESHOLDS, drwnDecisionTree::MIN_SAMPLES, and drwnDecisionTree::SPLIT_CRITERION for details.

See Also

drwnClassifier, drwnML Tutorial

---

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

• drwnDecisionTree.h
• drwnDecisionTree.cpp