Discriminant Analysis of Smart City Profiles Across Countries: A Multivariate Approach

According to the United Nations, more and more people will live in cities in the coming years. This poses serious challenges to local governments in terms of pressure on transport infrastructure, waste management, lighting, sewerage systems and the environment. In such a context, city management needs to become increasingly efficient and intelligent, thanks to the fundamental support of technology, the Internet of Things (IoT) and artificial intelligence (AI). The so-called Smart Cities (SCs) are thus developing across the globe to address this profound change. This study applies linear discriminant analysis (LDA) to a dataset of 102 smart cities from around the world, each described by six key dimensions of smartness: mobility, environment, government, economy, people, and living. The aim of this paper is to assess whether smart cities from the same country tend to share similar profiles and to identify the most distinctive national patterns in smart city development. Results show high accuracy in classification, with most of the cities correctly attributed to their country based on their smart characteristics. Canonical discriminant functions yield appreciable separation between countries, highlighting stable national models and outliers. Additionally, results suggest that national agendas and regional contexts play an important role in shaping smart cities.