Graphical method using neighborhoods for detecting outliers
We propose a new graphical method to help us to uncover potential outliers in multivariate samples. The idea behind the method is to analyze the behavior of a growing neighborhood of each data point. This method is very robust and allows to find outliers in very complex structures.
| Autor Principal: | Correa Morales, Juan Carlos; Universidad Nacional de Colombia |
|---|---|
| Otros Autores: | López Ríos, Víctor Ignacio; Universidad Nacional de Colombia |
| Formato: | info:eu-repo/semantics/article |
| Idioma: | eng |
| Publicado: |
Universidad Santo Tomás
2015
|
| Acceso en línea: |
http://revistas.usta.edu.co/index.php/estadistica/article/view/1430 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: |
We propose a new graphical method to help us to uncover potential outliers in multivariate samples. The idea behind the method is to analyze the behavior of a growing neighborhood of each data point. This method is very robust and allows to find outliers in very complex structures. |
|---|