Figure 2.

(a) SOM trained on spectra across the 5 eye tissue types. (b) SOMDI showing relative importance of different bands for each class to observed clustering in the SOM. (c) Classification accuracy of tissue using SKiNET against current state-of-the-art (multi-layer perceptrons (MLP), support vector machines (SVM), partial least squares discriminant analysis (PLS-DA) and k-nearest neighbours (kNN)). (d) Effect of number of principal components on classification accuracy for PCA based methods.