APA Class Reference

Averaged Passive-Aggressive Classifier. More...

#include <apa.h>

Inheritance diagram for APA:

Public Member Functions
void	train (Cases &cases, int T)
Y	predict (X &x)
virtual size_t	size ()
virtual void	save (std::ostream &os)=0
	Dump parameters to stream.
Public Attributes
unsigned	k
	k = \|Y\|
unsigned	d
	d = \|X\|
int	t
	t = current number of iterations
std::vector< std::string >	labels
Tanl::Text::WordIndex	predIndex
Float	margin
	Margin.
Static Public Attributes
static Float	C = 100.0
	aggressiveness
static Float	ka = 0.0
	kernel a<
static int	kd = 1
	kernel degree
static bool	verbose = false
	Control output.
static float	updatePercent = 0.5
Protected Member Functions
virtual void	update (Y y, Float tau, X &x)=0
virtual Float	score (unsigned i, X &v)=0
	Similarity score.
	APA (int k=0, int d=0)
virtual bool	load (std::istream &is)=0
	Load parameters from stream.
Static Protected Member Functions
static void	rand_permutation (std::vector< int > &OUT)

Detailed Description

Averaged Passive-Aggressive Classifier.

See also:: http://jmlr.csail.mit.edu/papers/volume7/crammer06a/crammer06a.pdf K. Crammer, O. Dekel, J. Keshet, S. Shalev-Shwartz. Online Passive-Aggressive Algorithms. Journ. of Machine Learning Research 7, 2006, 551-585.

virtual bool APA::load ( std::istream & is ) [protected, pure virtual]

Load parameters from stream.

Implemented in APAS, and APASV.

virtual size_t APA::size ( ) [inline, virtual]

Reimplemented in APAS, and APASV.

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