Forecasting Macro-Knowledge Competitiveness; Integrating Panel Data and Computational Intelligence

This paper is proposing an active framework for forecasting the competitiveness of knowledge based economy (KBE) for nations to serve the public, private, and policy makers. We have used three different forecasting methods including, panel data analysis, artificial neural networks (ANN) and linear multiple regression. By structuring and feeding balanced, panel data to the ANN framework, we were able to produce a better predicted results compared with the panel data fixed effect model and the linear multiple regression. The ANN framework can be applied in the context of forecasting the competitiveness of a KBE for any nation regardless of small time periods, or little data. To achieve this forecasting model, a two steps framework is proposed. The first step was to structure the data as panels, the second step is to create and train a three layers feed-forward ANN. The purpose of the first step was to exploit both the cross sectional and the time series variations on the same economy. The ANN was used to overcome the linearity of all types of regression models used and to create the trained framework which was capable of predicting the knowledge competitiveness and progress in any economy.