Sustainable development goals
  • Reduced inequality
Project link

About the project

SISBEN is a national survey that Colombia uses to classify and address the needs of its citizens. To classify the population according to socio-economic levels through the survey, a machine learning model was applied (Quentile Gradient Boosting), which presented excellent results in the evaluation of methodologies for SISBEN classification.

More project information

SISBEN is a survey that allows classifying the population according to their standards of living with the purpose of allowing social programs identify potential beneficiaries. The SISBEN IV methodology has the potential of allowing the identification of both income poverty and multidimensional poverty in the population.

The goal of the case is to classify the population in the best way possible as an input for the focalization of social programs considering possible biases in the reporting of information and out-of-date incentives.