No Detail Found!

COMPARISON OF DIFFERENT TECHNIQUES FOR DETECTION OF OUTLIERS IN CASE OF MULTIVARIATE DATA

Author(s)
Muhammad Zafar Iqbal Department of Mathematics and Statistics, University of Agriculture, Faisalabad, Pakistan
Samra Habib Department of Mathematics and Statistics, University of Agriculture, Faisalabad, Pakistan
Muhammad Imran Khan Department of Mathematics and Statistics, University of Agriculture, Faisalabad, Pakistan
Muhammad Kashif Department of Mathematics and Statistics, University of Agriculture, Faisalabad, Pakistan
Abstract
Yield trials play an important role for screening the competing cultivars. Often researchers normally measure several characteristics, commonly called as “parameters”, related to yield for each cultivar. Usually, researchers apply univariate statistical approaches even for multi-variable data; which limits the scope of the research implication. However, there is least tendency for considering multivariate techniques to analyze the multi-variable data. If results turn out weird than what are normally expected; then this leads the researcher to a dilemma. One of the possible reasons for such unexpected results is th e presence of outlier(s) in the data; which is often not obvious particularly in case of multivariate data. The focus of the present study is to provide the researchers with a comparative study of classical and robust techniques for the detection of outlier( s) where the data is recorded on multi variables. The statistical software R is used for the analysis. Results show that robust technique is more appropriate for detecting outlier(s) than the classical ones.
Publication Details
Page(s) 865-869
Published Journal: Pakistan Journal of Agricultural Sciences, Volume: 57, Issue: 3, Year: 2020

Request the full text from PASTIC

Keywords
outlier(s) detection, multivariate data, Screening cultivars, robust technique
0

Citations

0

Downloads

37

Views