Predicting and monitoring healthy diets in Rwanda with big data

Unhealthy diets are key to persistent problems in many low- and middle-income countries and are influenced by several external factors related to climatic, political and market conditions. Healthy foods are often less affordable, accessible and available. With the right data and analytics, it may be possible to understand and predict changes in diets over time, but current data is limited. Recently, a system for high-frequency monitoring of diet quality data was established in Rwanda, offering a unique opportunity to apply data science methods on a national scale, and develop an approach to enrich this data with data from other sources. The aim of this PhD research will be to use high-frequency self-reported data on diet quality in combination with other food system, climatic, agricultural and market related information, to identify the main determinants of a healthy diet and to achieve reliable predictions of the effects of nutrition interventions among rural and urban populations in Rwanda. The project forms part of an interdisciplinary programme on food security intelligence, in which a group of PhD researchers collaborate across different case studies that focus on the use of data to address food security challenges. The work will have a strong methodological focus and will target improvements in data collection processes, analytics, intervention design and impact evaluation.