Abstract

Artificial neural networks are computational structures inspired by biological neural systems. Conventional computational models are particularly well suited to executing sequences of instructions that have been precisely formulated for them. On the other hand, biological neural systems are well suited for tasks/operations such as speech, vision, information retrieval, generalization, and complex spatial and temporal pattern recognition in the presence of noisy distorted data, all of which are extremely difficult to accomplish by conventional computing methods. Therefore, the motivation for artificial neural networks (ANNs) is to achieve many of those desirable abilities of the biological neural systems.