Modelling (and forecasting) extremes in time series: A naive approach
2020
Neves, M.Manuela | Cordeiro, Clara
In Extreme Value Theory, we are essentially interested in the estimation of quantities related to extreme events. Whenever the focus is in large values, estimation is usually performed based on the largest k order statistics in the sample or on the excesses over a high level u. Here we are interested in modelling (and forecast- ing) extremes in time series. For modelling and forecasting classical time series, Boot.EXPOS is a computational procedure built in the R environment that has revealed to perform quite well in a large number of forecasting competitions. However, to deal with extreme values, a modification of that algorithm needs to be considered and is here under study
显示更多 [+] 显示较少 [-]