How does feedback in Artificial Neural Networks work?
The process of training a neural net to associate certain input patterns with correct output responses involves the use of repetitive examples and feedback. This iterative adjustment process continues for each hidden layer in the network until all neurodes have been subjected to the sensitivity analysis and the net is ready for another trial. 9. Briefly describe and compare supervised and unsupervised Artificial Neural Networks? The unsupervised nature of an unsupervised learning paradigm means that the ANN has been provided with the input data and examples but not with any feedback in the form of desired results. Using the clustering approach, a neural net can begin to “specialize” its learning with respect to a specific dimension of the data. The most common supervised learning paradigm is called back propagation. In this approach, the net receives an input example, generates a guess, and compares that guess to feedback containing the desired results. The ANN is being “supervised” by
