B0989
Title: funcharts: An R package for the real-time monitoring of multivariate functional data
Authors: Christian Capezza - University of Naples Federico II (Italy) [presenting]
Fabio Centofanti - University of Naples Federico II (Italy)
Antonio Lepore - Universita di Napoli Federico II (Italy)
Alessandra Menafoglio - Politecnico di Milano (Italy)
Biagio Palumbo - University of Naples Federico II (Italy)
Simone Vantini - Politecnico di Milano (Italy)
Abstract: The R package funcharts provides control charts for multivariate functional data and is based on the functional regression control chart framework introduced in the recent literature. This package is motivated by many modern industrial applications, where data are increasingly available as functional data or profiles. The quality characteristic of interest is often observed as a scalar or a functional quantity and is also affected by other variables, which in turn can be scalar variables or profiles. Functional regression models can be used to take into account the relationship between these variables. The main objective of funcharts is to provide appropriate methods for statistical process monitoring in these cases, with a particular focus on Phase II monitoring. Since in many industrial applications, functional data are observed over a temporal domain and it is important to give indications as soon as possible on possible anomalies, funcharts allows real-time monitoring, i.e., the monitoring of profiles possibly partially observed up to an intermediate domain point. A real-case study on monitoring CO2 emissions from navigation of a Ro-Pax cruise ship is shown to illustrate the practical applicability of the proposed R package.
