Predicting socio-economic status from satellite imagery using machine learning algorithm

This thesis explores the integration of satellite imagery and machine learning to predict socio-economic status (SES) across geographic regions. Traditional SES data collection methods, like surveys and censuses, are costly and infrequent. This study demonstrates how Convolutional Neural Networks (CNNs) provide a scalable, costecient alternative for estimating SES based on landscape features visible in satellite images.