Projects

Studying rural household economic shifts from agriculture to non-farm activities using large-scale survey data and econometric analysis. This project explores income, productivity, and poverty reduction pathways.

Developing advanced machine learning methods for identifying informative data features, moving beyond traditional methods like LASSO, Random Forest, and Mutual Information toward faster, more interpretable, and reliable models.

Designing a novel tree architecture integrating new mathematical diversity measures with a flexible structure adapted for complex data, enabling enhanced decision-making across domains.

Building an air pollution early warning system for fine particulate matter in Rwandan urban centres, particularly Kigali. Uses machine learning to capture tropical conditions and urban-rural dynamics to support policy intervention and public health protection.