Self-Organizing Maps | Conclusions
A Self Organizing Map is a type of artificial neural network but instead of learning from error-correction learning like backpropagation with gradient decent, it trains by using competitive learning
The model produces a two-dimensional representation of a higher dimensional data set while preserving the data structure
The data set is separated into clusters of variables with similar values which can be visualized in a map
This projection allows variables with similar values to be clustered together
Self organizing maps make higher-dimensional data easier to visualize and analyse
To understand how training progresses, plot the quantization and the topographic error for each step
The quantization error measures the quality of the learning and is equal to the average difference of the input samples compared to its corresponding winning neurons
The topographical error measures the projection quality and assesses the number of the data samples having the first best matching unit and the second best matching unit that are not adjacent