Poverty Mapping in Africa
Data on economic livelihoods plays a key role in enabling emerging markets to grow and in targeting philanthropic endeavors. In diverse and data scarce regions, such as Africa, acquiring this data can be a challenge. This talk centers on an AI-based poverty mapping system that is able to fill the data gaps. Leveraging the ever increasing availability of satellite imagery, the system is able to generate highly resolved poverty estimates for both the present and the past. Generated poverty estimates achieve an average 0.7 squared correlation against held out country data.
Anthony is a Machine Learning Engineer at AtlasAI, a small tech startup funded by the Rockefeller Foundation that generates actionable intelligence on agricultural and economic trends across the developing world. He graduated with an M.S. in Computer Science from Stanford University in 2018. At Stanford University he was a member of the Sustainability and Artificial Intelligence Lab. His is interested in deep learning and sustainability.