CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: Robust estimation in mixture models under case-wise and cell-wise contamination Authors:  Claudio Agostinelli - University of Trento (Italy) [presenting]
Giovanni Saraceno - University of Trento (Italy)
Ayanendranath Basu - Indian Statistical Institute (India)
Abstract: Classical contamination models consider as possible outliers the statistical units, that is, the whole observations (case-wise contamination). This approach has limits when it is applied to data sets with a large number of variables. A more general approach to robustness is to consider the possible presence of both case-wise and cell-wise contamination. The last, also known as the independent contamination model, identified as possible outliers in the cells. One approach is to filter out the contaminated cells from the data set and then apply a robust procedure able to handle case-wise outliers and missing values. We introduce filters in any dimension based on statistical data depth functions, and we show how we can use them in the robust estimation of parameters in a finite mixture model.