Towards Assessment of Innovativeness Economy Determinant Correlation: the Double Self-Organizing Feature Map Approach

Issue title: Dedicated to the Memory of Professor Manfred Kudlek

Article type: Research Article

Authors: Czyżewska, Marta | Szkoła, Jarosław | Pancerz, Krzysztof

Affiliations: University of Information Technology and Management, Sucharskiego Str. 2, 35-225 Rzeszów, Poland. [email protected], [email protected], [email protected]

Note: [] Address for correspondence: University of Information Technology and Management, Sucharskiego Str. 2, 35-225 Rzeszów, Poland

Abstract: In the paper, we attempt to identify the crucial determinants of innovativeness economy and the correlations between the determinants. We based our research on the Innovativeness Union Scoreboard (IUS) dataset. In order to solve the problem, we propose to use the Double Self-Organizing Feature Map (SOM) approach. In the first step, countries, described by determinants of innovativeness economy, are clustered using SOMs according to five year time series for each determinant separately. In the second step, results of the first step are clustered again using SOM to obtain the final correlation represented in the form of a minimal spanning tree. We propose some modifications of the clustering process using SOMs to improve classification results and efficiency of the learning process.

DOI: 10.3233/FI-2014-959

Journal: Fundamenta Informaticae, vol. 129, no. 1-2, pp. 37-48, 2014