Awesome
ROCK-PCA
ROtated Complex Kernel PCA for Spatio-Temporal Analysis of Earth Observation Data
This method is a conglomerate of different well know PCA methods traditionally used in geosciences as EOF. The purpose is to find spatio-temporal patterns in a unsupervised way, that represents the main modes of variability of geophysical variables, as described in:
D. Bueso, M. Piles and G. Camps-Valls, "Nonlinear PCA for Spatio-Temporal Analysis of Earth Observation Data," in IEEE Transactions on Geoscience and Remote Sensing. doi: 10.1109/TGRS.2020.2969813
https://ieeexplore.ieee.org/document/8989964
Folder:
- ROCK-PCA: contains the main ROCK-PCA code.
- demo: contains a easy-to-use toy example to test and try the method.
Please, don't have any doubt to contact us for suggestions and remarks.
Diego Bueso, Phd student, Universitat de València