| dc.description.abstract | Machine Learning (ML) is a rapidly growing subfield of Artificial Intelligence (AI) that involves the development of computer programs that learn by finding patterns in data, without being explicitly programmed. It has been successfully applied in various fields, including agriculture, manufacturing, marketing, finance, and healthcare. In the health sector, ML has been identified as having potential in drug development, clinical diagnosis, disease surveillance, outbreak response, and health systems management, by researchers and development partners such as WHO and USAID. Although ML has been touted for its potential to improve health outcomes, it has been used to a minimal degree in the DHIS2 ecosystem. Furthermore, recent developments have shown that some private organizations are carrying out the implementations. DHIS2 is a free and open-source digital platform that is used to collect and manage aggregated and patient health data, widely used in developing countries. The opportunities and challenges of using ML in health and developing countries have been studied. Still, none have specifically focused on how to integrate ML into digital health platforms like DHIS2 or the integration opportunities and challenges that developing countries would face where the platform is predominantly used. To investigate how ML can be integrated with DHIS2 and related opportunities and challenges, I conducted an interview-based study with various stakeholders from the DHIS2 community and used thematic analysis to analyze the data. The results suggest that ML can be integrated with the platform through an app that can be implemented in both standalone and client-server architectures. The opportunities included improved forecasting, disease predictions, and anomaly detections. The challenges included the need for large amounts of data, data quality problems, lack of experts, lack of awareness, inadequate supporting infrastructure, insufficient funding, maintenance costs, risks of project failures, and lack of guidelines for the use and development of ML applications. | eng |