The New and Computationally Efficient MIL-SOM Algorithm: Potential Benefits for Visualization and Analysis of a Large-Scale High-Dimensional Clinically Acquired Geographic Data

Figure 4

illustrates 𝑈-Matrices and clusters derived from the standard SOM ((a), (c), (e), and (g)) and MIL-SOM ((b), (d), (f), and (h)) algorithms. Experimental datasets include childhood asthma ((a), (b)); adult asthma ((c), (d)); computer random generated ((e), (f)); and blood lead levels ((g), (h)). Map size 40×40 neurons and the topology of the neurons are hexagonally in shape.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)