CIRFA welcomes you to a seminar presented by CIRFA PhD student Katalin Blix on the topic:
Machine Learning Gaussian Process Regression for Remote Sensing Applications
The powerful Gaussian Process Regression (GPR) method has been shown to perform excellently in comparison to other machine learning techniques and parametric approaches. GPR has several advantageous properties besides its regression strength, such as the certainty level of the estimates and the possibility to retrieve information about the relevance of features.
We give a brief introduction to the principles of the GPR approach and show how the method can be used for different remote sensing practises. The application areas we present here include water quality parameter estimation from optical imaging data in various aquatic environments and sea ice remote sensing.