Weakly supervised method to predict economic development: A case study for Nepal and Cambodia

Abstract

Reliable and timely measurements of economic activities are critical for understanding economic development and helpful in delivering humanitarian aid and disaster relief where needed. However, many developing countries still lack reliable data. This paper introduces a novel human-machine collaborative algorithm for measuring eco- nomic development from high-resolution satellite images in the absence of ground truth statistics. The novelty of our method is that we break down a computationally challenging problem into sub-tasks, which involves a human-in-the-loop solution that incor- porates lightweight human annotations on economic development in the machine-learning process. We demonstrate how to apply our method to developing country economies (e.g., Nepal and Cam- bodia) with insufficient data and provide reliable and inexpensive indicators of economic development on a granular level.