Time series outlier detection: a new non parametric methodology (washer)

VENTURINI, ANDREA; Divisione Analisi E Ricerca Economica Territoriale, Banca D’Italia – Sede Di Venezia
The production and exploitation of statistical data for a large amount of high frequency time series must allow a timely use of data ensuring a minimum quality standard. This work provides a new outlier detection methodology (washer): efficient for timesaving elaboration and implementation procedures, adaptable for general assumptions and for needing very short time series, reliable and effective as involving robust non parametric test. Some simulations, a case study and a ready-to-use R-language function (washer.AV())...