Tag: poverty alleviation

Rating of relative poverty/welfare level

If you have experience from research on the African continent, take the opportunity to participate in this survey by researchers from Lund University.


Predict is part of a larger project which is being undertaken by a group of researchers at the Department of Human Geography, Lund University. The study aims at measuring the level of welfare/poverty in villages from satellite data.


While poverty measurement has, conventionally, been undertaken using nationally-representative surveys such as the Demographic and Health Survey, the downsides to using such traditional methods of welfare measurement are well-publicized. In more recent times, new methods in machine learning, especially in the specific area of deep learning, have been applied on very high-resolution satellite data to estimate welfare levels with increasingly improving accuracy. An interesting question then is whether machine learning algorithms would perform better than or with human experts at predicting poverty levels from imagery.

The results from this project will provide valuable insights into what can be deduced from visual observations of landscapes in terms of welfare and answer the question of whether a combination of experts and AI can better predict poverty.

Researchers and contact information

Ola Hall, Head of Department and Associate professor, Department of Human Geography at Lund University

Ibrahim Wahab, Postdoctoral fellow, Department of Human Geography at Lund University