Land degradation and climate change are increasingly converging to threaten ecosystem stability, food security, and rural livelihoods in China, making land restoration a critical pathway for strengthening socio-ecological resilience. However, many existing land restoration decision-support studies still use simple linear methods, consider ecological, governance, and social factors separately, and do not properly include community resilience and local knowledge under uncertainty, which makes it difficult to select the most suitable restoration strategies. This study proposes an integrated decision-support framework that combines the analytic network process (ANP), artificial neural networks (ANN), and fuzzy weighted aggregated sum product assessment (fuzzy WASPAS) to prioritize restoration strategies in China. Four main criteria and 16 sub-criteria were developed in this study. ANP results indicate that ecological stabilization and climate resilience drivers dominate the weighting structure, with vegetation recovery and land-cover stabilization, soil erosion control, drought tolerance, and long-term ecosystem stability emerging as the most influential sub-criteria, while policy alignment and institutional support represent key enabling conditions for implementation. ANN refinement improved the robustness of the ANP-derived weights by optimizing the distribution of closely competing factors and reducing subjectivity-driven variability, providing a more stable priority structure for the fuzzy WASPAS strategy evaluation. Using the final ANP-ANN weights, fuzzy WASPAS results rank the seven restoration strategies, showing that the hybrid co-governance restoration model achieves the highest overall utility, followed by climate-smart adaptive restoration and community-led stewardship.