CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: Statistical data depth aimed for text data Authors:  Alicia Nieto-Reyes - Universidad de Cantabria (Spain) [presenting]
Abstract: Text data has particular characteristics when transformed into quantitative data. In particular, it can be high dimensional with each datum having many zero-value components. Generally, statistical data depth functions (multivariate or functional) do not perform well under this scenario. Thus, some transformations of these functions will be proposed to tackle the problem. The performance will be shown through an application to health care text data.