Odor discrimination using adaptive resonance theory

Author(s): Distante, C. | Siciliano, P. | Vasanelli, L. |

Year: 2000

Citation: Sensors and Actuators B: Chemical. Volume 69, Issue 3, 248-252

Abstract: The paper presents two neural networks based on the adaptive resonance theory (ART) for the recognition of several odors subjected to drift. The neural networks developed by Grossberg (supervised and unsupervised) have been used for two different drift behaviors. One in which the clusters end up to overlap each other and the other when they do not. The latter case is solved by unsupervision, which is useful to track the moving clusters and possibly discover new odors autonomously.

Topics: Machine Learning, Applications: Chemical Analysis, Models: ART 2 / Fuzzy ART, Fuzzy ARTMAP,

PDF download

Cross References