Note: This project is for CS 591.003 "Data Mining" and CS 591.001 "Big data". This is not meant to be used for predictive purposes and any use of this project is at your own risk. There are no guarantees for the validity of any results used by this program.
The Growing Neural Gas (GNG) network is a type of neural network that is dynamic throughout time. The idea behind the alogorithm is to generate a neural network that can form to any data by 'diffusing' through the Euclidean space. Upon each iteration, the GNG will slowly dift closer to the underlying data, approximating its structure. These neurons will form and break edges as the local error of each nueron decreases. It terms of clustering, as the network removes edges, components will begin to separate and disconnected components can be considered "cluters."
More to come...