Predicting student outcomes with machine learning

20180726_141224 (1).jpg
This August, Theta led a machine learning hackathon with flexible online learning service TANZ eCampus. This involved a mixed team of nine participants including IT and business stakeholders from TANZ, along with Theta Analytics consultants Peter George, Konrad Jarocinski and Willem Pelser.
The hackathon used machine learning to help predict student outcomes for participants in online courses offered through TANZ eCampus.  Learner outcomes are of great importance to TANZ, and providing early indicators of students who may need additional support is valuable.
The event was great fun – the TANZ team got a fresh perspective on what is possible with machine learning, and the Theta team came away with an increased appreciation for the environment that TANZ operates in.
Shane Wohlers, Learning Analytics Manager for TANZ eCampus, sums it up:
We were searching for a hands-on introduction to machine learning and predictive modelling, with an aim to start building our in-house capability--Theta’s workshop was exactly what we had hoped for.
Theta took the time to understand our goals and what we wanted to achieve. They worked with us to articulate the problems we wanted to explore and ensure we had the right data, people, skills and business insights there in the room. The two days were focused on understanding and applying machine learning to solve those problems.
I was impressed with Theta’s level of knowledge and insight and how open and accessible they were in sharing that with us. They took the time to demystify and explain key principles, strategies, and scenarios relevant to us and work alongside our staff to create a suite of functioning prototypes. We left feeling excited, having taken our first steps on a new path.