David Goody is a data scientist and operational researcher at the Department for Education. David will speak on how the Department for Education provides data-driven oversight for over 5000 education providers and ensures the regularity and propriety of funds spent within them. This involves using a scaleable systematic process of data collection and analysis to help identify and intervene where appropriate in cases of financial concern. His work has included developing predictive risk analysis systems, geo-spatial analysis tools, automated impact assessment modelling tools and performance measures. He has provided support for data science development across government and won the best speaker award at the 2016 Government Operational Research Service conference.
Key OR techniques that are used to support this work are:
- Clustering and machine learning techniques to allow us to review and analyse a wide range of large and dynamic data sources. These include decision tree learning, lasso regression, k-means clustering and k nearest neighbouring clustering.
- Advanced visualisations, including using interactive geo-spatial analysis showing areas served by different education providers.
- Combining multiple data sources to create effective MI reports and presenting these in ways that non-analysts can understand.
In the afternoon, David will speak on 'How to Make the Leap into Data Science', with examples of data science techniques in action from the department’s work and examples that audience members can replicate.
The Annual Analytics Summit takes place 21 June 2016, London.