11:50am - 12:30pm Predictions, based on machine learning models, change the approach for development of the university
Tuesday Jun 11
For many years the university has used annual descriptive data for major strategic decisions. Typical the collection of official benchmark indicators. Now we has decided to act more data-driven with a focus on using predictions based on machine learning models. Predictions gives the management new possibilities to act in time regarding promoting potentials or reducing risks at the most important key performance indicators. Then the decision processes aren’t no longer “same procedure as last year?” – now the data challenges us to rethink the approach to how we manage and develop the university. Student behavior can be predicted with benefits for both the students and the university. For example, we can predict 8 of 10 student drop-outs, and we can predict 6 of 10 students who are delayed during their studies. These kind of results challenges the university's performance and strategy. But it also opens for new actions and possibilities.