Self-Organizing Maps and Their Applications in Image Processing, Information Organization, and Retrieval - Nonlinear Signal and Image Processing: Theory, Methods and Applications - page 413

1.6

opt
SOM
MMSOM 1.4

1.2

1

0.8

0.6

0.4

0.2

0

0.2 0.3 0.4 0.5 0.6 0.7 0.8

epsilon

(a)

3

opt
SOM
MMSOM

2.5

2

1.5

1

0.5

0.2 0.3 0.4 0.5 0.6 0.7 0.8

epsilon

(b)

FIGURE 11.6
Bias introduced by the SOM, MMSOM, and the nearest mean reclassification algorithm ( T opt )in
estimating (a) the true mean for the dominating cluster; (b) the true means for both clusters vs.
the contamination percentage for σ = 3.