Layman's tutorial to train and fit an SOM, explained in detail at: algobeans.com/self-organizing-map
Iterative changes to SOM grid visualized for conceptual understanding:
Number of iterations taken for neurons in the SOM grid to stabilize:
Frequency map of neurons in the SOM grid:
U-matrix showing similarity and dissimilarity netween neurons. Useful for identifying superclusters:
Radar plots showing the aggregated profile of each neuron, as determined by their constituent data points:
Size of neuron labels is based on the number of corresponding data points it is associated with:
Layman's tutorial to train and fit an SOM, explained in detail at: algobeans.com/self-organizing-map