class StumpTrainer

This class is designed to train a StumpMachine for a 2-class problem (positive patterns/negative patterns) using histograms

Inheritance:


Public Fields

[more]bool verbose
[more]StumpMachine* stump_machine
the weak classifier to train
[more]int bestFeature
id of the best feature
[more]real bestThreshold
threshold of the best features
[more]int bestDirection
direction of the best feature
[more]int n_pos
number of positive patterns
[more]real histPos[HIST_PRECISION_]
histogram of positive patterns
[more]real hCumulPos[HIST_PRECISION_]
cumulative histogram of positive patterns
[more]int n_neg
number of negative patterns
[more]real histNeg[HIST_PRECISION_]
histogram of negative patterns
[more]real hCumulNeg[HIST_PRECISION_]
cumulative histogram of negative patterns

Public Methods

[more]virtual void train(DataSet* data_, MeasurerList* measurers)
[more]virtual ~StumpTrainer()

Public Members

[more] -----


Documentation

This class is designed to train a StumpMachine for a 2-class problem (positive patterns/negative patterns) using histograms

obool verbose

oStumpMachine* stump_machine
the weak classifier to train

oint bestFeature
id of the best feature

oreal bestThreshold
threshold of the best features

oint bestDirection
direction of the best feature

oint n_pos
number of positive patterns

oreal histPos[HIST_PRECISION_]
histogram of positive patterns

oreal hCumulPos[HIST_PRECISION_]
cumulative histogram of positive patterns

oint n_neg
number of negative patterns

oreal histNeg[HIST_PRECISION_]
histogram of negative patterns

oreal hCumulNeg[HIST_PRECISION_]
cumulative histogram of negative patterns

o -----
-----

ovirtual void train(DataSet* data_, MeasurerList* measurers)

ovirtual ~StumpTrainer()


This class has no child classes.
Author:
Yann Rodriguez (yann.rodriguez@idiap.ch) Sebastien Marcel (marcel@idiap.ch)
Version:
2.0 Fri 15 Jul 2005 11:35:27 AM CEST
Since:
1.0

Alphabetic index HTML hierarchy of classes or Java



This page was generated with the help of DOC++.