Towards a Smart Industrial Policy: Identifying Leading and Lagging Sectors in Saudi Arabia’s Vision 2030 Framework Using Machine Learning

Saudi Arabia’s Vision 2030 establishes economic diversification and reduced dependence on oil revenues as national priorities, with industrial development forming a central point of this transformation. An effective industrial policy, however, requires the ability to distinguish between leading, neutral and lagging sectors using reliable and timely evidence. Traditional approaches often rely on expert judgement or static surveys, which may lack the responsiveness required in dynamic economic environments. This study introduces a data-driven framework that integrates machine learning and deep learning techniques to classify and forecast sectoral performance based on the Industrial Production Index. The methodology combines weighted multi-feature clustering and rule-based adjustments to enhance interpretability, while long shortterm memory (LSTM) networks were employed to forecast short-term directions. Results demonstrated a possibility of classifying sectors into leading, neutral and lagging categories, with activities such as electrical equipment and chemicals showing robust growth, whereas basic metals and furniture exhibited sustained contraction. Visual analytics complements these findings by providing decision-makers with accessible tools to monitor industrial dynamics. The study underscores the value of data-driven approaches in informing targeted interventions, supporting industrial upgrading and guiding diversification policies under Vision 2030.