A1224
Title: On boxplots for functional data
Authors: Stanislav Nagy - Charles University (Czechia) [presenting]
Tomas Mrkvicka - University of South Bohemia (Czech Republic)
Antonio Elias - Universidad de Malaga (Spain)
Abstract: The functional boxplot is a useful visualization tool for datasets where each observation is a function. Such boxplots are based on the notion of functional depth, ranking observations from typical to more extreme. While in principle any functional depth can be used to construct a boxplot, the most common depths in practice are integrated depths, e.g., the popular modified band depth. However, integrated depths do not result in well-defined boxplots. Instead, infimal depths are the only functional depths that result in a valid construction of a functional boxplot.
