What is a Competitive Network?
A competitive network is typically a type of unsupervised machine learning that uses competitive learning to deliver results. Through specific mathematical and network modeling, competing networks achieve different goals in the recognition and processing of inputs.
Competing networks are also known as competing neural networks.
The key to the design of competing networks is the idea of excitatory and inhibitory influences on an artificial neuron. Experts speak of lateral inhibition and feedback connections through which laterally placed nodes can inhibit others. At the same time, forward excitatory synaptic connections can provide other appropriate influences. Understanding how these work together is part of the roadmap to working closely with competitive networks for new outcomes in machine learning and artificial intelligence.