Bipropagation
From EverybodyWiki Bios & Wiki
Bipropagation is a machine-learning algorithm invented in 2009 by Bojan Ploj. It is intended for feed-forward neural nets like a multi-layer perceptron. With this algorithm, supervised learning is done which is appropriate for both kinds of neural network usage – for classification and for regression. Bipropagation is an improvement of famous Backpropagation and is much faster than it and more reliable. With bipropagation, the XOR problem is solved 20 times faster and each attempt is successful, which is not the case with backpropagation.
