Strategic Educational Planning Through Deep Learning: A 1D-CNN Forecasting Model for SDG 4

Ghada Alturif¹Alaa Abdelbary²Radwa Ahmed Osman Mohamed^2*

• ¹Princess Nourah bint Abdulrahman University, P.O. Box 1029, Riyadh 11671, Saudi Arabia, Saudi Arabia
• ²Arab Academy for Science, Technology and Maritime Transport (AASTMT), Alexandria, Egypt

The paper uses a time series forecasting approach to anticipate educational achievements for the United States, Saudi Arabia, China, Egypt, and Sweden (2025-2030), supporting SDG 4's aims of inclusive and excellent education. A one-dimensional Convolutional Neural Network (1D-CNN) is used to analyze socioeconomic, demographic, and educational data from 2000 to 2022, capturing temporal patterns across key indicators like enrollment rates, literacy levels, teacher-student ratios, and digital access in learning environments. The 1D-CNN architecture was chosen because of its shown ability to handle sequential data and detect minor changes in educational markers. The model is trained on validated historical data and delivers country-specific projections that are consistent with observed educational growth trends. The results show that China and Sweden are predicted to attain near-complete SDG 4 compliance, with scores nearing 100\%, while the United States is expected to maintain high levels over 92\%. Saudi Arabia makes consistent growth, with estimated values ranging from 78 to 80\%, while Egypt is expected to have the lowest results, ranging from 65 to 68\%. These findings demonstrate the model's potential to generate accurate and interpretable projections over a wide range of educational variables, providing useful insights for policymakers, educators, and international organizations. By delivering data-driven evidence, the strategy facilitates focused resource allocation and strategic planning, hence accelerating progress toward SDG 4.