Examinando por Materia "Mahalanobis distance"
Mostrando 1 - 3 de 3
Resultados por página
Opciones de ordenación
Ítem Estimación robusta de la matriz de covarianza para la selección óptima de portafolios de inversión(UNIV NAC COLOMBIA, FAC NAC MINAS, 2018-01-01) Gutiérrez-Sepúlveda D.; Laniado H.; Medina-Hurtado S.; Universidad EAFIT. Escuela de Ciencias; Modelado MatemáticoThe selection of portfolios under the Media-Variance (M-V) model work bad when it is exposed to the presence of atypical data that generate error estimation of the parameters In order to minimize this estimation error, we investigate new robust methodologies and their financial performance in terms off the ratio Sharpe, of the turnover index and of the variance. The estimation of the covariance matrix parameter is done with three different robust methods that seek to minimize the instability generated by atypical data, the first is the great contribution of this research, which consists in shrinking the covariance matrix with a cut-out to the mean, the second and third methods are chi-square cut-outs in the distance of Mahalanobis and Minimum Determinant of the Covariance Matrix (MCD) respectively. © The author; licensee Universidad Nacional de Colombia.Ítem Modelos de clasificación basados en ingeniería de características robusta para la predicción de deserción de clientes en el sector financiero(Universidad EAFIT, 2024) Quiroz Fino, Jefferson Adolfo; Ortiz Arias, SantiagoÍtem Outliers in semi-parametric Estimation of Treatment Effects(Universidad EAFIT, 2017-10-30) Ugarte Ontiveros, Darwin; Canavire-Bacarreza, Gustavo; Castro Peñarrieta, Luis; gcanavir@eafit.edu.coAverage treatment effects estimands can present significant bias under the presence of outliers. Moreover, outliers can be particularly hard to detect, creating bias and inconsistency in the semi-parametric ATE estimads. In this paper, we use Monte Carlo simulations to demonstrate that semi-parametric methods, such as matching, are biased in the presence of outliers. Bad and good leverage points outliers are considered. The bias arises because bad leverage points completely change the distribution of the metrics used to define counterfactuals. Whereas good leverage points increase the chance of breaking the common support condition and distort the balance of the covariates and which may push practitioners to misspecify the propensity score. We provide some clues to diagnose the presence of outliers and propose a reweighting estimator that is robust against outliers based on the Stahel-Donoho multivariate estimator of scale and location. An application of this estimator to LaLonde (1986) data allows us to explain the Dehejia and Wahba (2002) and Smith and Todd (2005) debate on the inability of matching estimators to deal with the evaluation problem.