Table 8 Classification results of Salinas Scene with 50 dimensions: Data with actual dimensions, reduced through PCA, ICA, vanilla AE, and D-Path-AE are used to train three classifiers; Decision Tree, K-Nearest Neighbours, and Support Vector Machine. Kappa, OA, and AA are the three measures, that demonstrate the classification accuracy of reduced data.

From: Enhancing feature learning of hyperspectral imaging using shallow autoencoder by adding parallel paths encoding