A Hybrid Forecasting Model for Spanish Unemployment: How COVID-19 Destroys Statistics
64th ISI World Statistics Congress - Ottawa, Canada
Format: CPS Abstract
Session: CPS 34 - Impact of covid IV
Monday 17 July 5:30 p.m. - 6:30 p.m. (Canada/Eastern)
Unemployment prediction is a cornerstone of a national economic planning, thus an adequate forecasting model becomes urgent. The nonlinear nature of the data set requires advanced techniques that are able to capture these nonlinearities. Moreover, due to the regional characteristics of the Spanish labor market, the ability to capture the seasonal patterns is desirable. In the present paper, a hybrid model based on a linear model with neural network correction is considered. The results of the simulations show that the proposed model gives an accurate forecasting, however, the COVID-19 makes the further prediction impossible. As with any other crisis, it provokes an unpredictable behavior and the collapse of seasonal patterns.