JuGAAD Framework for Downscaling Socioeconomic Indicators via Autoencoders
July 24, 2026
JuGAAD uses an autoencoder to compress high-dimensional NSSO survey data into low-dimensional latent representations, which are then mapped to fine-grained census and geospatial data. This three-step framework enables the downscaling of coarse socioeconomic indicators to village-cluster scales in India.
HOW THIS AFFECTS YOU
●
researcherYou can use this multi-step regression and latent mapping approach to bridge scale mismatches in sparse socioeconomic datasets.