CPerceptron Class Reference

---

Detailed Description

Class Perceptron implements the standard linear (online) perceptron.

Given a maximum number of iterations (the standard perceptron algorithm is not guaranteed to converge) and a fixed lerning rate, the result is a linear classifier.

See also:

CLinearClassifier

http://en.wikipedia.org/wiki/Perceptron

Definition at line 30 of file Perceptron.h.

Inheritance diagram for CPerceptron:

List of all members.

Public Member Functions
	CPerceptron ()
	CPerceptron (CDotFeatures *traindat, CLabels *trainlab)
virtual	~CPerceptron ()
virtual EClassifierType	get_classifier_type ()
virtual bool	train (CFeatures *data=NULL)
void	set_learn_rate (float64_t r)
	set learn rate of gradient descent training algorithm
void	set_max_iter (int32_t i)
	set maximum number of iterations
virtual const char *	get_name () const
Protected Attributes
float64_t	learn_rate
int32_t	max_iter

---

Constructor & Destructor Documentation

CPerceptron

(

)

default constructor

Definition at line 17 of file Perceptron.cpp.

CPerceptron	(	CDotFeatures *	traindat,
		CLabels *	trainlab
	)

constructor

Parameters:

	traindat	training features
	trainlab	labels for training features

Definition at line 22 of file Perceptron.cpp.

~CPerceptron

(

)

[virtual]

Definition at line 29 of file Perceptron.cpp.

---

Member Function Documentation

virtual EClassifierType get_classifier_type

(

)

[virtual]

get classifier type

Returns:

classifier type PERCEPTRON

Reimplemented from CClassifier.

Definition at line 48 of file Perceptron.h.

virtual const char* get_name

(

void

)

const [virtual]

Returns:

object name

Reimplemented from CLinearClassifier.

Definition at line 73 of file Perceptron.h.

void set_learn_rate

(

float64_t

r

)

set learn rate of gradient descent training algorithm

Definition at line 61 of file Perceptron.h.

void set_max_iter

(

int32_t

i

)

set maximum number of iterations

Definition at line 67 of file Perceptron.h.

bool train

(

CFeatures *

data = NULL

)

[virtual]

train classifier

Parameters:

data

training data (parameter can be avoided if distance or kernel-based classifiers are used and distance/kernels are initialized with train data)

Returns:

whether training was successful

Reimplemented from CClassifier.

Definition at line 33 of file Perceptron.cpp.

---

Member Data Documentation

float64_t learn_rate [protected]

learning rate

Definition at line 77 of file Perceptron.h.

int32_t max_iter [protected]

maximum number of iterations

Definition at line 79 of file Perceptron.h.

---

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

• Perceptron.h
• Perceptron.cpp