Skip to main content
. 2020 Sep 30;12:31. doi: 10.3389/fnsyn.2020.00031

Figure 2.

Figure 2

Application of principal component analysis (PCA) method to dendritic spine clusterization. Spine shape is characterized by a set of parameters (A). PCA method is used to reduce data dimensionality before clusterization. Newly generated parameters called principal components composed from initial one and form an orthonormal basis (B). After PCA dataset in new coordinates is subjected to clusterization (C). Cluster shape and content depend on the clusterization method used (k-means, c-means, hierarchical, et cetera).