6533b85efe1ef96bd12bf448
RESEARCH PRODUCT
Large-scale random features for kernel regression
Diego Marcos GonzalezValero LaparraGustau Camps-vallsDevis Tuiasubject
Computer science1900 General Earth and Planetary Sciencescomputer.software_genreKernel (linear algebra)10122 Institute of GeographyKernel methodWavelet1706 Computer Science ApplicationsRadiative transferLife ScienceKernel regressionData mining910 Geography & travelcomputerdescription
Kernel methods constitute a family of powerful machine learning algorithms, which have found wide use in remote sensing and geosciences. However, kernel methods are still not widely adopted because of the high computational cost when dealing with large scale problems, such as the inversion of radiative transfer models. This paper introduces the method of random kitchen sinks (RKS) for fast statistical retrieval of bio-geo-physical parameters. The RKS method allows to approximate a kernel matrix with a set of random bases sampled from the Fourier domain. We extend their use to other bases, such as wavelets, stumps, and Walsh expansions. We show that kernel regression is now possible for datasets with millions of examples and high dimensionality. Examples on atmospheric parameter retrieval from infrared sounders and biophysical parameter retrieval by inverting PROSAIL radiative transfer models with simulated Sentinel-2 data show the effectiveness of the technique.
year | journal | country | edition | language |
---|---|---|---|---|
2015-07-31 | 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) |