Fuzzy Modeling for HLA Typing

G.B. Ferrara, F. Masulli, C. Pera, S. Rovetta, R. Sensi

Ottavo Convegno dell'Associazione Italiana per l'Intelligenza Artificiale
Siena, 10-13 Settembre 2002

Abstract

Oligonucleotide microarrays are able to perform a large quantity of simultaneous
experiments, resulting in a digital image to be analyzed. We describe the image
analysis step of a system designed for HLA typing based on microarray technology
and employing an adaptive fuzzy system which can learn from user hints. The fuzzy
modeling approach allows using the life scientist’s language and concepts to describe
and classify the probe activations in a natural way, and allows robust interactive image
filtering thanks to the adaptive behaviour of the fuzzy system.