Binary tree genetic algorithm with Quadtree for land cover classifications

Author

Ng, Sai-Cheong ; Leung, Kwong-Sak

Author_Institution

Dept. of Comput. Sci. & Eng., Chinese Univ. of Hong Kong

Volume

5

fYear

2004

fDate

20-24 Sept. 2004

Firstpage

3050

Abstract

Enhanced binary tree genetic algorithm (BTGA+) has been successfully applied to land cover classification problems. However, the execution time of BTGA+ is quite long on large datasets. In this paper, a novel decision tree algorithm, called binary tree genetic algorithm with Quadtree (BTGA with quadtree), is proposed by extending BTGA+. In the proposed algorithm, a generalized Quadtree is constructed when a new node of a linear decision tree is created. The proposed algorithm runs faster than BTGA+ on datasets with sufficiently large number of samples, without sacrificing the quality of decision trees constructed by BTGA+

Keywords

decision trees; genetic algorithms; image classification; quadtrees; terrain mapping; vegetation mapping; BTGA+; binary tree genetic algorithm; decision tree algorithm; land cover classifications; linear decision tree; quadtree; Binary trees; Biological cells; Classification tree analysis; Computer science; Decision trees; Genetic algorithms; Genetic engineering; Genetic mutations; Impurities; Tree data structures;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004 IEEE International

Conference_Location

Anchorage, AK

Print_ISBN

0-7803-8742-2

Type

conf

DOI

10.1109/IGARSS.2004.1370341

Filename

1370341