Multi-agent systems for Classification of E-Nose data

Abstract

Metal Oxide Semiconductor Gas Sensors are used to measure and classify odors. This kind of system requires both advanced sensor design and classification techniques. In this paper we present a MOGS (Metal Oxide Gas Sensor) specifically designed to classify the breath of humans. We propose an architecture that incorporates new sensing technology and a classification technique based on multi-agent systems. The proposal is evaluated using samples from Asian and European participants. The results obtained are promising.