Predicting household socioeconomic position in Mozambique using satellite and household imagery

TL;DR

This study demonstrates that combining satellite and ground-based household imagery with machine learning can effectively predict household socioeconomic status at the individual level in Mozambique, offering a scalable approach for socioeconomic assessment.

Contribution

It introduces a novel approach using multimodal imagery and explainable AI to predict household SEP at the household level, expanding beyond previous area-level predictions.

Findings

Asset-based SEP prediction achieved highest accuracy with all image types.

Ground-based photographs help focus on relevant household elements for SEP prediction.

The reduced model with key household elements performs nearly as well as the full model.

Abstract

Many studies have predicted SocioEconomic Position (SEP) for aggregated spatial units such as villages using satellite data, but SEP prediction at the household level and other sources of imagery have not been yet explored. We assembled a dataset of 975 households in a semi-rural district in southern Mozambique, consisting of self-reported asset, expenditure, and income SEP data, as well as multimodal imagery including satellite images and a ground-based photograph survey of 11 household elements. We fine-tuned a convolutional neural network to extract feature vectors from the images, which we then used in regression analyzes to model household SEP using different sets of image types. The best prediction performance was found when modeling asset-based SEP using random forest models with all image types, while the performance for expenditure- and income-based SEP was lower. Using SHAP, we observed clear differences between the images with the largest positive and negative effects, as well as identified the most relevant household elements in the predictions. Finally, we fitted an additional reduced model using only the identified relevant household elements, which had an only slightly lower performance compared to models using all images. Our results show how ground-based household photographs allow to zoom in from an area-level to an individual household prediction while minimizing the data collection effort by using explainable machine learning. The developed workflow can be potentially integrated into routine household surveys, where the collected household imagery could be used for other purposes, such as refined asset characterization and environmental exposure assessment.