Patricia McAlernon, Jonathan M. Slater and King-Tong Lau
An optimized gas sensor array consisting of eight quartz microbalance sensors coated with different sorbent layers was used in conjunction with a Kohonen network to discriminate between different organic vapours. The sensing materials used in the array were chosen to promote the range of different solvation interactions possible with the sample test set. The feature extraction capabilities of the Kohonen network were realized in that Kohonen output patterns were more easily distinguished than the original sensor response patterns. The Kohonen network was also shown to preserve relationships in the data.