Munich Personal RePEc Archive

Forecasting the role of public expenditure in economic growth Using DEA-neural network approach

Amiri, Arshia and Ventelou, Bruno (2011): Forecasting the role of public expenditure in economic growth Using DEA-neural network approach.

[img]
Preview
PDF
MPRA_paper_33955.pdf

Download (98Kb) | Preview

Abstract

This paper integrates data envelopment analysis (DEA) and artificial neural networks (ANN) to forecast the role of public expenditure in economic growth in OCDE countries. The results show that this approach is a powerful and appropriate method to forecast this role. DEA method allows us to develop a neutral evaluation, unbiased a priori by any type of criteria, of the proportions in which the goal of productive spending is pursued, for any expenditure. Then we apply ANN to forecast economic growth by using input data taken at frontier. At the end of the DEA-ANN chain, prediction-power tests appear positive: best structures of multiple hidden layers indicate more ability to forecast according to best structures of single hidden layer but the difference between those is not much.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.