PERFORMANCE OF CLUSTERING ALGORITHMS AND CLUSTER VALIDITY INDICES ON HYPERSPECTRAL DATA

منو

" PERFORMANCE OF CLUSTERING ALGORITHMS AND CLUSTER VALIDITY INDICES ON HYPERSPECTRAL DATA "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	150557
Doc. No	:	ET22349
Main Entry	:	Felix M. Fonth
Title Proper	:	PERFORMANCE OF CLUSTERING ALGORITHMS AND CLUSTER VALIDITY INDICES ON HYPERSPECTRAL DATA
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	Hyperspectral such as AVIRIS and HYDICE can collect image data in up to 224 and 21 0spectral narrow bands respectively, where each spectral band represents a dimension onthe Feature Space. Very often, validated data of these images is not available for trainingsupervised classifiers, and not even the number of possible classes in the data set isknown, but one is presented with the task of classifying the data points in the image.This work presents the performance of three different unsupervised classificationalgorithms on Hyperspectral data. These algorithms are the Fuzzy C-Means (FCM)Clustering algorithm, the Kohonen Clustering Network (KCN) algorithm, and the FuzzyKohonen Clustering Network (FKCN) algorithm,.-...,..tested for theQ1 PC1 bus cardBoth these projects mere sofixare des elopment efforts tonards contributing to dlfferentaspects of Roboucs and lZ1echatronics projects m the Controls and Roboucs Group..
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL45555.pdf
Title and statement of responsibility and	:	PERFORMANCE OF CLUSTERING ALGORITHMS AND CLUSTER VALIDITY INDICES ON HYPERSPECTRAL DATA [Thesis]