Skip to content

Latest commit

 

History

History
8 lines (5 loc) · 503 Bytes

README.md

File metadata and controls

8 lines (5 loc) · 503 Bytes

Self Organizing Map (Kohonen Map)

The package is based on MiniSom toolbox, with some modification by (oliviaguest).

I am using these maps to model infant learning in a dual categorization task.

The SOM class extends the standard SOM with noisy encoding capability, as well as a decision rule. It assumes that category presentation denotes category membership.

The SOM2 class treats the side feature as an extra layer that is associated with the BMU (based on just object feature information).