The project was aimed towards aiding municipalities of South Africa to better measure their performance in terms of Service Delivery and Efficiency. We achieved this through the creation of a profiling measure. Furthermore we investigated which youth characteristics are associated with youth unemployment. An extensive tool was developed in the form of a dashboard, where the results and real time predictions of the developed models can be viewed.
- We used several machine learning techiques:
- Principal Componet Analysis (PCA): to find the features which affect youth employment.
- Decision Tree and Random Forests : to predictions the probability of youth getting employed given the characteristics of their municipality.
- Support Vector Machines (SVM) : to predict the profile of the municipality.
- Outputs to the projects:
- Here is the Link to the data source: