World Data Lab specializes in creating economic and demographic models that help organizations make data-driven decisions. Our global team of data scientists, economists, demographers, and developers are passionate about combining new data sources with groundbreaking methods to push the frontiers of what is knowable.
As technology evolves, so do our methods for modeling and forecasting poverty. Since 2018, World Data Lab has been experimenting with machine learning as a mechanism for processing daylight satellite images as well as uncovering powerful relationships between geographic features and economic activity. This method offers the promise of mapping entire countries and continents at the 5-10 square kilometer level using a consistent and reliable approach. There is little doubt that AI-based poverty mapping will soon fundamentally revolutionize the way public planning and development activities are designed and dellivered.
The Convolutional Neural Network (CNN) model produces highly granular poverty estimates for selected geographies by using a method that combines the latest developments in computer vision and deep learning with data sourced from daylight, nightlight, and geo-coded survey microdata. This method requires a wide range of technologies to process the data and develop the full cycle of modeling, including: AI frameworks, database frameworks, geometry and satellite data processing tools, clouds, virtual machines, multiple programming languages, and graphical processing units.
Partner with us to apply machine learning techniques to your poverty mapping project. In addition to providing custom code and outputs from our AI-based algorithm, we also provide training sessions regarding the method and the tools needed to deliver it.