PROJECTION PURSUIT ALGORITHMS TO DETECT OUTLIERS

Ciudad de Mexico

Workshop; EUREKA 2015 Fifth international workshop on Knowledge Discovery, Knowledge Management and Decision Support; 2015

Eureka Iberoamerica and Eureka International networks, the Mexican Society of Operations Research, and Universidad Autónoma Metropolitana (UAM)

An important heterogeneity between observations, canbe presented in the presence of outliers. Which is whythey have been detected. Besides they distort measuresdescriptive and parameters population estimates.There are different multivariate methods to detectoutliers, among which distance-based methods andprojection pursuit methods. The last don´t distribution ofdata assumptions, but require high computational load.In this paper, we compare the results of the applicationof methods proposed by Peña and Prieto 2001 andFiltzmoser et. al. 2007 to a group of Argentinecompanies.