Quantifying prediction uncertainty for functional-and-scalar to functional autoregressive models under shape constraints (Q1733283)

Demand and supply time series forecasting in energy markets is of course a hot topic both in energy economics and engineering, see, e.g. [\textit{D. N. Sidorov} et al., Izv. Irkutsk. Gos. Univ., Ser. Mat. 26, 76--90 (2018; Zbl 1409.45003)]. This paper proposes the novel moving block bootstrap procedure generating several bootstrap forecasts which gives useful indications as to the possible trajectory of the future curve. Point-wise confidence intervals can be plotted from these trajectories. Taking into account such bootstrap trajectories, the resulting forecasting model can benefit from the ensembling of many predictions, similarly to how random forest works, see [\textit{A. V. Zhukov} and \textit{D. N. Sidorov}, Vestn. Yuzhno-Ural. Gos. Univ., Ser. Mat. Model. Program. 9, No. 4, 86--95 (2016; Zbl 1442.68210)].