WEAKLEARNER |
Produce classifier thresholding single feature.
Synopsis:
model = weaklearner(data)
Description:
This function produce a weak binary classifier which assigns
input vector x to classes [1,2] based on thresholding a single
feature. The output is a model which defines the threshold
and feature index such that the weighted error is minimized.
This weak learner can be used with the AdaBoost classifier
(see 'help adaboost') as a feature selection method.
Input:
data [struct] Training data:
.X [dim x num_data] Training vectors.
.y [1 x num_data] Binary labels (1 or 2).
.D [1 x num_data] Weights of training vectors (optional).
If not given then D is set to be uniform distribution.
Output:
model [struct] Binary linear classifier:
.W [dim x 1] Normal vector of hyperplane.
.b [1x1] Bias of the hyperplane.
.fun = 'linclass'.
Example:
help adaboost
See also:
ADABOOST, ADACLASS.
About: Statistical Pattern Recognition Toolbox
(C) 1999-2004, Written by Vojtech Franc and Vaclav Hlavac
Czech Technical University Prague
Faculty of Electrical Engineering
Center for Machine Perception
Modifications:
25-aug-2004, VF
11-aug-2004, VF