Self-organizing Map
A self-organizing map (SOM) or self-organizing feature map (SOFM) is an unsupervised machine learning technique used to produce a low-dimensional (typically two-dimensional) representation of a higher dimensional data set while preserving the topological structure of the data.
How to use:
- Upload an image - it will be used as the training data
- Adjust learning parameters to control how the map adapts
- Watch as the neural network organizes itself to match the image
Parameters explained:
- Initial Range: How far neurons influence their neighbors initially
- Learning Rate: How quickly neurons adapt to the data
- Decay Rates: How these values decrease over time
- Weight factors: Balance between position and color matching
The SOM is a type of artificial neural network that creates a 2D map of high-dimensional input. It's used for dimensionality reduction, visualization, and clustering. Watch as the network self-organizes to represent the color and spatial structure of your image.