Abstract

The purpose of this research study is two-fold. First to evaluate and compare the innovativeness of Greece relative to the European Union using indicators from the European innovation scoreboard and second to propose practices and techniques concerning the utilization of machine learning for modeling and analyzing innovation in general. Systematic analysis is conducted regarding the over-performance and the under-performance of Greece and the trends of these indicators over the years through statistical techniques and methods. Machine learning and advanced statistical methods are incorporated to ascertain the most important features that drive the variation of the summary innovation score of the European Union and Greece. Clusters and groups of correlated indicators are also specified. Our study provides preliminary explanations and evidence to help the country value its advantages and deal with its disadvantages. It also paves the way towards an analysis of innovation approach that has the potential to help us elucidate certain issues and deeply understand observations concerning Greece and beyond.

Keywords: Innovation, Data science, Data analysis, Machine learning, Statistics