Figure 6
Example for transferring reference c-layers to a target data set. Panels (A,B) represent the results of the supervised and unsupervised learning of the target data set. Panels (C,D) show the probability of the assignment of the reference c-layers to the target c-layers as determined by an SVM ensemble and the final discrimination result. Unsupervised clustering was based on feature set F1, and the supervised learning on F3. The dataset under discrimination is the RSC area M262L0, and the reference data for supervised learning are the c-layers of the 8th hierarchical level of VISp area M248L1, that is, the one shown in Fig. 4. For comparison purposes, the manually labelled layers are indicated to the right of panel (D). Panel (E) shows a corresponding representative x–z plane, and the bottom panels six x-y plane slice examples at z = 50 µm (F), z = 90 µm (G), z = 140 µm (H), z = 220 µm (I) z = 300 µm (J), and z = 400 µm (K), corresponding to the six c-layers on the 7th hierarchical level in panel B. Both panels (E,F) correspond to the first d-layer in panel (D), whereas panels (G–I) correspond to the second d-layer, and panel J corresponds to the third d-layer.
