Self-Organizing Maps
SOMs is a technique that takes multidimensional data and reducing it to one or two dimensions. This compression technique is called vector quantization. The technique usually involves a visual component that allows a human to better see how the data has been categorized. SOM learns without supervision.
The SOM is good for finding clusters, which is not to be confused with classification. With classification we are interested in finding the best fit for a data instance among predefined categories. With clustering we are interested in grouping instances where the categories are unknown.
A SOM uses a lattice of neurons, usually a two-dimensional array or a hexagonal grid, representing neurons that are assigned weights. The ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access