We believe that …
We can implement AI and Machine Learning approaches to support the delivery of Leadership Academy national programmes, including better targeting of programmes and personalisation of offers, which would improve participant experiences and outcomes. It would also enable us to gain richer insights from programme-related data that we collect and thus enhance our evaluations, allowing us to continually improve and widen access to our programmes.
How critical is this hypothesis?
We aim to use this research to …
Inform the introduction of data-driven course matching for participants based on their individual characteristics and evaluation of groups who take up our national programme offers, and demonstrate how supervised and unsupervised machine learning algorithms to automate data analysis can be deployed in Leadership Academy programmes as a proof of concept.
To verify that, we will…
- Develop a classification algorithm (supervised learning) to predict the most suitable course for a participant using their demographic and employment information.
- Develop an unsupervised clustering model to identify the features of distinct groups of individuals who take up Leadership Academy programmes, and discover who we currently do and do not serve.
- Identify opportunities to scale up machine learning applications in Leadership and Lifelong Learning work and embed these in national programme delivery for real-time use.
- The accuracy of the classification algorithm
- The separability in feature space between distinct clusters of participants
We are right if…
We can correctly predict participants’ course choices based on their demographics and characterise the features of distinct groups of participants.
Non-urgent advice: Experiment #003
Subject: AI / ML development
Emma Mi, Ella Mi
30 Sept 2020
1 Oct 2020
Based on the Strategizer lean test card