Abstract | ||
---|---|---|
In this work, we face the problem of training sample collection for the estimation of biophysical parameters by adopting the active learning approach. In particular, we propose two active learning strategies specifically developed for Gaussian Process (GP) regression. The first one is based on adding samples that are distant from the current training samples in the kernel space while the second one exploits an intrinsic GP regression outcome to pick up the most difficult samples. Experiments on simulated and real data sets show the effectiveness of active selection of training samples for regression problems. |
Year | DOI | Venue |
---|---|---|
2011 | 10.1109/IGARSS.2011.6049994 | Geoscience and Remote Sensing Symposium |
Keywords | Field | DocType |
Gaussian processes,biological techniques,learning (artificial intelligence),medical computing,parameter estimation,regression analysis,remote sensing,GP regression,Gaussian process regression,active learning approach,biophysical parameter estimation,kernel space,training sample collection problem,Active learning,Gaussian process (GP) regression.,biophysical parameters,chlorophyll concentration estimation | Kernel (linear algebra),Kriging,Data set,Active learning,Pattern recognition,Regression,Computer science,Regression analysis,Gaussian process,Artificial intelligence,Estimation theory,Machine learning | Conference |
ISSN | ISBN | Citations |
2153-6996 | 978-1-4577-1003-2 | 4 |
PageRank | References | Authors |
0.59 | 11 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Edoardo Pasolli | 1 | 285 | 17.04 |
Farid Melgani | 2 | 1100 | 80.98 |