A New Learning Algorithm for the Fuzzy Adaptive Resonance Theory: Multispectral Classification of the Algiers’s Bay

This paper presents a new learning algorithm for the fuzzy adaptive resonance theory. The modification allows us to supervise the fuzzy ART and to simplify ARTMAP network. It consists to find network’s parameters (comparison, training and vigilance) which gave the minimum quadratic distances between the output of the training base and those obtained by the network. A comparative study of these two parameterized network and an third modified fuzzy ARTMAP are done. In this last network, learning is done differently. We don’t take account of the eight (08) values of network’s parameters. As application we carried out a classification of the image of Algiers’s bay taken by SPOT XS. The results of this study presented in the forms of curves, tables and images show that modified fuzzy ARTMAP presents the best compromise quality/computing time.