Projection pursuit algorithms to detect outliers

Ciudad de México

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)

A significant heterogeneity between observations can be a consequence of the presence of outliers. The detection of outliers is an important task for the statistical analysis since they distort descriptive measures and parameters estimators.There are different multivariate methods to detect outliers, such as distance-based methods and projection pursuit methods. In this paper, we compare the methods proposed by Peña and Prieto (2001) and Filtzmoser et al. (2007) to detect outliers in a set of Argentine companies that quote their shares in the Stock Exchange.