Artificial Neural Networks Predict Sustainable Development Goals Index

The Sustainable Development Goals Index (SDGI) serves as a critical measure of global progress toward achieving the United Nations' 17 Sustainable Development Goals (SDGs), encompassing environmental, social, and economic dimensions. However, the SDGI’s utility is challenged by the high dimensionality and complexity arising from over 200 indicators, limiting transparency and practical application for policymakers and researchers. This study introduces a novel, efficient artificial intelligence (AI)-driven approach combining artificial neural networks (ANNs) and genetic algorithms (GAs) to dramatically simplify SDGI prediction while maintaining remarkable accuracy. Leveraging data extracted from seven authoritative global reports, including the Environmental Performance Index, Global Innovation Index, and Human Development Index, a total of 288 operational sustainability indicators covering 127 countries were initially considered. Through systematic collinearity analysis and performance-based feature selection, the indicator set was refined to 90 effective predictors. Employing a GA-optimized ANN framework, our model identifies a minimal set of four crucial indicators—deaths from infectious diseases, ICT usage, education expenditure, and academic assessment scores—that predict the SDGI with an accuracy exceeding 97% (correlation coefficient 0.96). The findings underscore the pivotal role of innovation-related metrics and unveil a surprising marginalization of environmental indicators in the current SDGI construct, calling into question potential biases in environmental sustainability assessment. This parsimonious yet powerful predictive model not only facilitates streamlined SDG monitoring but also offers fresh insights into the influencer variables driving sustainable development worldwide. These results hold profound implications for global sustainability measurement practices by enabling more transparent, accessible, and actionable SDG evaluations. The synergy of advanced machine learning techniques and comprehensive global data creates a scalable framework for future research on sustainable development analytics, policy simulation, and real-time SDG tracking. Keywords: Sustainable Development Goals Index, Artificial Neural Networks, Genetic Algorithms, Feature Selection, Innovation, Environmental Bias, Global Sustainability Indicators, Machine Learning, Predictive Modeling, Sustainable Development Analytics