Extreme Learning Machine for Remote Sensing Classification

Abstract

Extreme Learning Machine for Remote Sensing Classification

Mahesh Pal
Assistant Professor
Department of Civil Engineering, NIT,
India
Email: mpce_pal@yahoo.co.uk

This paper explores the potential of extreme learning machine based supervised classification algorithm for land cover classification. In comparison to a backpropagation neural network, which requires setting of several user-defined parameters and may produce local minima, extreme learning machine require setting of one parameter and produce a unique solution. ETM+ multispectral data set (England) was used to judge the suitability of extreme learning machine for remote sensing classifications. A back propagation neural network was used to compare its performance in term of classification accuracy and computational cost. A classification accuracy of 89% is achieved by using extreme learning machine with this dataset in comparison to an accuracy of 87.75% provided by a backpropagation neural network. Thus, suggesting a better performance by extreme learning machine for this dataset. A comparison of computational cost suggests a better performance by extreme learning machine (1.25 second) in comparison to back propagation neural network (336.20 seconds).