Author(s)

F. Sorbello, A. Tarantino, M. Tarantino & G. Vassallo

Abstract

In this work it will be described how the artificial neural networks and multilayer perceptrons in particular, can be successfully employed to solve two real world tasks such as the character recognition and the identification of the functional groups in organic compounds. In the first it will be underlined the requirement to have a simple architecture with good performance (high speed and recognition rates) in order to implement it in a chip. It will be described the choices that have allowed to obtain these targets beginning from the pre-processing technique and continuing with th

Keywords