Introduction
The aims of the application is to provide a tool for experimenting and modeling
neural networks.
An artificial neuron is an automaton and a neural net can be seen as a net of
interconnected automata.
To define a network the connection architecture and
the kind of automata are to be given: that is the network structure and the updating rule, to specify the
propagating patterns or activities. Moreover a network can adapt itself to inputs, modifying connections weights. This process is called the learning process.
Neural networks provide an effective approach to variuos sets of problems, varying from
pattern recognition, or similar problems involving patterns, to images analysis or robot
control. Many interesting applications in different fields involve pattern analysis and
classification: speech recognition, medical images, automatic hand writing treatment and
so on. These are the areas in which neural networks have shown to be most promising.
Back to index