UK schools analysis with agglomerative clustering

3D scatter plot showing UK schools values for teacher headcount, number of pupils on free meal and IDACI score.

This weekend I wanted to learn a little bit more about Plotly Dash. I wanted to build something quickly with the idea of using real-world data.

So, I come up with a simple cluster analysis of some of the best data sets openly available on the Web: UK schools performances.

This datasets it’s a gold mine of information about schools in UK including:

• School census data,
• Financials of the school for a particular year,
• Pupils results on various stages
• Workforce

I also wanted to join this data set with the UK Deprivation Index to show potential correlations between the economics of the catchment area of a school and that school features.

I then ran an agglomerative cluster algorithm to spot affinities between schools and I’ve assembled everything up in a Plotly dashboard.

The full demo is available here.

Technologies used: Pandas, Scikit-learn, Plotly. All assembled with ❤️on SherlockML.