Leveraging national surveys, big-data advances and machine learning, Cornell researchers have piloted a new approach to mapping poverty that could help policymakers identify the neediest people in poor countries and target resources more effectively.