Abstract—Climate change poses a major challenge to resource-dependent communities, particularly in developing countries where agriculture remains the primary livelihood. This study analyzes farmers' perceptions of climate change impacts and their adaptation strategies based on a survey of 385 farmers conducted in Oulad Said commune, Settat Province (Morocco), a representative semi-arid region of North Africa. Using a Multivariate Probit (MVP) model framed within an AI-enhanced machine learning perspective, the analysis captures interdependencies among multiple adaptation decisions while integrating predictive validation. The results indicate that climate change has significantly reduced agricultural production and rural livelihoods. Farmers' adaptation choices are positively influenced by socioeconomic factors such as income, access to credit, education, and geographical origin, whereas larger household size and livestock ownership are associated with a lower likelihood of adaptation, reflecting resource constraints and limited flexibility. By embedding the MVP model within a hybrid econometric–machine learning framework, this study demonstrates the value of combining interpretability and predictive capacity to better understand farmers' behavioral responses to climate risks. The findings offer policy-relevant insights for promoting climate-resilient and sustainable agricultural systems in Morocco and other climate-vulnerable regions
Keywords—climate change adaptation, artificial intelligence, machine learning, multivariate probit model, predictive analytics, sustainable resource systems, agricultural resilience
Cite: Amine Hmid, Redouane Kaiss, Abdelghafour Achy, Youssef Laababid, and Aziz Boutaieb, "AI-driven Machine Learning for Farmers' Climate Change Adaptation Using the Multivariate Probit Model: Insights for Sustainable Resource Systems," International Journal of Environmental Science and Development vol. 17, no. 3, pp. 201-215, 2026.
Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).