A Study on the Factors of Revolutionary Destabilization in Asian and African Countries Using Machine Learning Methods Has Been Published

A team of researchers from the Center for Interdisciplinary Studies of Revolutionary Processes — Ivan Chernomorchenko, Ilya Medvedev, and Andrey Korotayev — has published an article in the journal Cross-Cultural Research (ranked at the List A) titled Towards the Ranking of the Importance of Revolutionary Destabilization Factors in Asian and African Countries Using Machine Learning Methods. The study focuses on identifying the key structural drivers of armed and unarmed revolutionary destabilization in Sub-Saharan Africa, the Middle East and North Africa, and Asia over the period from 1950 to 2022.

In contrast to traditional statistical approaches, the authors employed interpretable machine learning models (the CatBoost gradient boosting algorithm combined with SHAP analysis), which allowed them to uncover nonlinear and region‑specific effects that standard regression methods tend to miss.

The methodological novelty of the work lies in combining the high predictive power of machine learning with post‑hoc interpretability (SHAP values and permutation importance), thereby bridging the gap between accurate forecasting and theoretically meaningful explanation.

The authors conclude that no universal “pan‑continental” explanations of revolutionary destabilization exist. Armed and unarmed trajectories represent qualitatively distinct phenomena with their own causal architectures, which vary systematically across Asia, Africa, and the Middle East.