Identifying Outliers - A Look By A Grand Tour
Price
Free (open access)
Transaction
Volume
30
Pages
9
Published
1998
Size
676 kb
Paper DOI
10.2495/ASE980021
Copyright
WIT Press
Author(s)
Anna Bartkowiak
Abstract
Applied Sciences and the Environment 13 Identifying outliers - a look by a grand tour Anna Bartkowiak EMail: aba@ii.uni.wroc.pl Abstract The paper considers mainly a graphical method, called the grand tour which permits to 'see' multivariate data points located in more then 3- dimensional space. The essence of the method is to make a quasi-continuous sequence of projections from the considered multivariate space. The outlying (atypical) points may then be seen at outstanding positions. The method is illustrated when considering some official air-pollution data: Emissions of pollutants (Sulfur and Nitrous Dioxides) in 1985 and in 1989 by 24 European countries. Four countries were doubtlessly identified by the grand tour as outstanding in their pollutant emissions: When con- sidering the total amount of the emission (in thousands of tons), these were: Federal Republic of Germany, German Democratic Republic, United Ki
Keywords