A new comparative model for national innovation systems based on machine learning classification techniques

Dados Bibliográficos

ID6149901
AutoresIbrahim Alnafrah (0000-0002-7448-8315, ITMO University, autor correspondente), Bassel Zeno (0000-0002-7820-8210, ITMO University)
Ano2019
Volume10
Fascículo1
Páginas45-66
Data de publicação2019-01-11
Peer ReviewedSim
Open AccessNão
TipoARTICLE
PeriódicoInnovation and Development (JOURNAL)
Identificadores do periódicoISSN: 2157-9318 • E-ISSN: 2157-930X
EditoraTaylor & Francis (PUBLISHER • GB)
DOI10.1080/2157930x.2018.1564124
OpenAlexW2908922857
IdiomaEN
Citações recebidas3
Referências citadas32

This study aims to cluster and classify national innovation systems (NISs) dynamically based on analysing the structural differences among NISs’ dimensions. This study provides a tool that will help policymakers monitor the process of building and development NIS.Regarding the methodology, machine learning classification and clustering techniques were used, in which clusters represent three level of development: high, medium and low NISs’ clusters.The empirical study includes 36 indicators from 54 countries over 29 years (1980–2008), which are divided into six groups, that represent the different NISs’ dimensions.The results of clustering show a high level of similarity between clusters and the economic and innovation reality in studied countries. Moreover, the results of classification models indicate a high level of accuracy. These models are considered a good tool for monitoring the development process of NIS and enabling policymakers to improve their innovation strategies to accelerate NIS’s development process

Cluster (spacecraft · Cluster analysis · Cluster development · Data mining · Machine learning · Process (computing · Similarity (geometry · Computer Science · Energy, Environment, Economic Growth · Engineering · Firm Innovation and Growth · International Business and FDI · Artificial Intelligence