What is it about?
The main aim of this research is to analyse the impact of the forecast horizon on model selection. To this end, we use data on international tourism demand for all seventeen regions of Spain. The study assesses the influence of the forecast horizon on the predictive performance of several machine learning techniques. We compare the accuracy of Support Vector Regression (SVR) models to that of Neural Network (NN) models, using a linear time-series model as a benchmark.
Featured Image
Photo by Philip Veater on Unsplash
Why is it important?
We find that the SVR model with a Gaussian radial basis function kernel outperforms the rest of the models for the longest forecast horizons. We also find that machine learning methods improve their forecasting accuracy with respect to linear models as forecast horizons increase.
Perspectives
The obtained results show the suitability of SVR models for medium and long-term forecasting.
Oscar Claveria
AQR-IREA, Univeristy of Barcelona
Read the Original
This page is a summary of: Regional Forecasting with Support Vector Regressions: The Case of Spain, SSRN Electronic Journal, January 2015, Elsevier,
DOI: 10.2139/ssrn.2945533.
You can read the full text:
Resources
Contributors
The following have contributed to this page
