Abstract:
Support vector machines (SVMs) have been developed during the last two decades and recently acknowledged as very effective methods for general purpose pattern recognition...View moreMetadata
Abstract:
Support vector machines (SVMs) have been developed during the last two decades and recently acknowledged as very effective methods for general purpose pattern recognition. The important key in using a SVM is to select the appropriate parameters of its kernel function. In this paper, we present techniques on adjusting kernel parameters of SVMs to improve their performances with two remote sensing texture classification problems.
Date of Conference: 20-24 September 2004
Date Added to IEEE Xplore: 27 December 2004
Print ISBN:0-7803-8742-2