Title: Functional profile completion: An R package Authors:  Matus Maciak - Charles University (Czech Republic) [presenting]
Abstract: Risk assessment is typically based on various parametric approaches using rather strict theoretical assumptions, but real data often do not comply with these restrictions. Alternatively, non-restrictive and easily applicable nonparametric functional-based techniques relying on completion of missing profiles can be used. We present the first R package that fully implements such nonparametric risk assessment techniques. Three competitive algorithms are discussed in particular. A brief theoretical background and important statistical properties are addressed together with some practical illustrations and real datasets.