ClassifierThreshold< N > Class Template Reference

#include <ClassifierThreshold.h>

Public Member Functions
void	Update (const Feature &feature, int target)
int	Classify (const Feature &feature) const
void	Reset ()
void *	GetDistribution (int target) const

Detailed Description

template<int N>
class ClassifierThreshold< N >

This class use N-D Gaussian distribution for pos and neg data set. Given a feature, it returns whether it's positive or negative.

Author

Zhengrong Wang.

Member Function Documentation

template<int N>

int ClassifierThreshold< N >::Classify

(

const Feature &

feature

)

const

Classify the feature. Use the simple Euclidean distance to pos and neg cluster.

Parameters

feature

the feature extracted, should be N dimension

Returns

1 for positive, -1 for negative

template<int N>

void ClassifierThreshold< N >::Reset ( )

inline

Reset everything to the start state.

template<int N>

void ClassifierThreshold< N >::Update	(	const Feature &	feature,
		int	target
	)

Use Kalman filter to update the pos and neg cluster.

Parameters

feature	the feature extracted, should be N dimension.
target	positive > 0, negative <= 0.

---

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

• include/ClassifierThreshold.h