A Kohonen Self-Organising Feature Map - type of neural network that relies on unsupervised learning to extrapolate statistical patterns from input and train itself. The KSOFM is sensitive to differences in its input data and evokes its own classes based on these differences. Its output vector is an 2D array onto which these classes are mapped in clusters. Based on neurological evidence that the human brain works in a similar fashion.

Log in or register to write something here or to contact authors.