Application of Neural Network Models for Analysis of Factors Influencing Patent Activity

Application of Neural Network Models for Analysis of Factors Influencing Patent Activity

The purpose of this article is to present an idea for the application of Neural Network models (NNM) in the analysis of the relationship between certain types of resources allocated to innovation and patent activity. The European Union (EU) and Member States, as well as most countries around the world, are interested in investing in R&D to create and maintain competitive advantages. The question arises about the effectiveness of every euro spent on innovation. The article introduces a NNM that aims to determine the sensitivity of patent activity in certain EU countries to the dynamics of some input variables. Input variables reflect the resource spending for R&D – intramural R&D expenditure, total R&D personnel and researchers, number of scientists and engineers. Patent activity in countries is evaluated on the Number of patent applications to the European Patent Office (EPO) per year variable. As a result of model application, it is found that the sensitivity analyzed varies from country to country, but in any case, the higher the resource spending for R&D, the higher the patent activity is. The article draws a series of conclusions, the main one being that patent activity is most sensitive to changes in the variable R&D expenditure.