A. Elías, J. M. Morales González, S. Pineda Morente
In this work, we aim to use functional depths to study the temporal dynamic of time-dependent functional data. More concretely, we tackle the analysis of Functional Time Series (FTS) which are sets of sample curves indexed in time, and High Dimensional Functional Time Series (HDFTS) that terms the situation when several FTS are under analysis. With this goal, we explore different well-known time dependency structures (FARMA, SFAR, etc.), and we show how univariate time series of functional depths can retain key information of the functional temporal dependency. Then, we show how this time series can be exploited to study the temporal dependency of FTS and to detect evolution outliers in the context of HDFTS.
The two methodologies are empirically tested by simulation and their applicability is illustrated with actual smart-metering data corresponding to photo-voltaic energy generation and circuit voltage records.
Keywords: functional time series, functional depths, evolution outliers, smart meters data
Scheduled
GT06 Functional Data Analysis I. New methodologies
June 8, 2022 12:40 PM
Grade Hall