Distributed Extreme Learning Machine with Kernels Based on Mapreduce for Spectral-Spatial Classification of Hyperspectral Image

ELM with kernels and MapReduce have an unparalleled advantage of other similar technologies, which attract widely attention in machine learning and distributed data processing communities respectively. In this paper, we combine the advantage of ELM with kernels and MapReduce, and propose a Distributed Extreme Learning Machine with kernels based on MapReduce framework (DK-ELMM),which makes full use of the parallel computing ability of MapReduce framework and realizes efficient learning of large-scale training data. In particular, we present a spectral-spatial DELMM-based classifier for hyperspectral remote sensing images that integrates the information provided by extended morphological profiles. The proposed spectral-spatial classifier allows different weights for both (spatial and spectral) features outperforming other ELM-based classifiers in terms of accuracy for land cover applications. The accuracy classification results are also better than those obtained by equivalent spectral-spatial SVM-based classifiers.