GENLSDATA |
Generates linearly separable binary data.
Synopsis:
data = genlsdata(dim,num_data,margin)
Description:
It generates randomly binary labeled vectors which
are linearly separable with prescribed margin.
Input:
dim [1x1] Data dimension.
num_data [1x1] Number of generated data.
margin [1x1] Minimal ensured margin (distance of the closest
vector to the separating hyperplane).
Output:
data [struct] Generated data:
.X [dim x num_data] Sample data.
.y [1 x num_data] Data labels (1 or 2).
model [struct] Ground truth linear classifier:
.W [dim x 1] Normal vector of separating hyperplane.
.b [1x1] Bias of the hyperplane.
Example:
data = genlsdata(2,50,1);
model = ekozinec( data );
model.margin
figure; ppatterns(data); pline(model);
See also
PERCEPTRON, EKOZINEC, LINCLASS, SVM.
About: Statistical Pattern Recognition Toolbox
(C) 1999-2003, Written by Vojtech Franc and Vaclav Hlavac
Czech Technical University Prague
Faculty of Electrical Engineering
Center for Machine Perception
Modifications:
3-may-2004, VF
16-Feb-2003, VF
26-feb-2001 V.Franc