Ítem
| Universitat de Girona. Departament d’Informàtica i Matemàtica Aplicada | |
| Hijazi, R. | |
| Zeros in compositional data are classified into “rounded” zeros and “essential” zeros. Therounded zero corresponds to a small proportion or below detection limit value while the essentialzero is an indication of the complete absence of the component in the composition. Several parametricand non-parametric imputation techniques have been proposed to replace rounded zeros andmodel the essential zeros under logratio model. In this paper, a new method based on EM algorithmis proposed for replacing rounded zeros. The proposed method is illustrated using simulateddata | |
| http://hdl.handle.net/2072/273635 | |
| eng | |
| Universitat de Girona. Departament d’Informàtica i Matemàtica Aplicada | |
| Tots els drets reservats | |
|
Estadística matemàtica -- Congressos
Mathematical statistics -- Congresses Anàlisi multivariable -- Congressos Mathematical statistics -- Congresses |
|
| An EM-Algorithm Based Method to Deal with Rounded Zeros in Compositional Data under Dirichlet Models | |
| info:eu-repo/semantics/conferenceObject | |
| Recercat |
