When we estimate GDP for lower income countries, we usually have a constraint of both data quality and data sparsity. What should we do to solve such a problem? Machine learning techniques together with alternative data sources are used here to have a nowcasting model for lower income countries.
