1. Design and simulation of a multilayered hardware structure of a programmable Cellular Neural Network (a kind of locally connected network using feedback and clipping integrating characteristics) for high-resolution image processing based on amorphous or polycrystalline silicon thin-film technology, under the guidance of prof. Valerio Cimagalli and ing. Marco Balsi at TCE Lab. This project has evolved into a graduation dissertation titled "Simulazione di circuiti neurali in silicio amorfo e policristallino" (7/14/97).

2. A Perceptron simulator running under Matlab environment with different learning algorithms (by now the classical back propagation and the one step secant, a quasi Newton method); it will become the core of a general package on perceptrons.

I'd like to suggest a very good introduction on neural networks from the University of Amsterdam.

Neural Network Information Page