Company or Institution

University of Pretoria

Industry

Education

Website

https://dsfsi.github.io/blog/seminar-saair/

Country

South Africa

Sustainable Development Goals (SDGs)

General description of the AI solution

Higher education institutions face ramifications such as budget cuts related to the high dropout rates experienced globally in institutions of higher learning. As a result, research in higher education has concluded a variety of areas that require improvement based on traditional university practices. One of these areas is the improvement and scalability of student support strategies. Student support strategies is a collective term used to overarch processes within the institution that if implemented correctly, have an impact on student throughput and retention rates. These include student tutorials, academic advising, writing interventions and other social support provided by units within the institution. The biggest problem with these strategies is in both the timing of them, and the manpower required to scale them. As a result, this study explores the depth of creating a generalized framework to build systems within higher education institutions by using a combination of machine learning and reinforcement learning. The proposed framework is applicable to generalized datasets and can be adopted in a variety of different contexts. This study proved how such a framework should be implemented for higher education institutions.

Github, open data repository, prototype or working demo

https://github.com/dsfsi/Higher_Education_EDA

Publications

Combrink, H., & Oosthuizen, L. (2022). Strengthening online teaching and learning by closing the feedback loop. Southern African Review of Education with Education with Production, 27(1), 57-76.

Herkulaas, M. and Oosthuizen, L.L., 2020. First-Year Student Transition at the University of the Free State during COVID-19: Challenges and Insights. Journal of Student Affairs in Africa, 8(2), pp.31-44.

Combrink, H. M. V. E., van der Merwe, N. C., Katarya, R., de Wet, K., & Motloung, M. H. (2022). A South African Indian population group dataset for breast cancer and BRCA1/2 variants. Data in Brief, 42, 108180.

Needs