Global trends in nuclear energy forecasting: Methodological analysis, bibliometric evidence and strategic-geopolitical dimensions

The aim of this study is to reveal global trends in the production and consumption of nuclear energy. It will also compare econometric, machine learning and hybrid models using common criteria. Furthermore, it will evaluate bibliometric findings within a strategic-geopolitical context. The study covers 36 research articles published between 1 August 1996 and 11 November 2025. Data were collected from the Web of Science Core Collection (WoSCC), Dimensions and PubMed databases, and analyses were performed using VOSviewer and Biblioshiny software. 66.7% of the publications were produced after 2011, a period which accounted for 96.5% of the total citations. Overall, 1,117 citations and an h-index of 17 were obtained. On a country basis, the highest production (15 publications) belongs to China. The findings also demonstrate that, since 2011, machine learning-based approaches, such as Support Vector Regression (SVR) and Long Short-Term Memory (LSTM), have emerged as dominant, progressing in parallel with the energy security and decarbonisation agenda. This study provides an evidence-based framework for selecting methods and planning policies for forecasting nuclear energy.