Predictive analytics

Published By GOV.UK [English], Mon, Sep 26, 2022 4:22 AM


The growing capacity of the data science work at the Government Actuary’s Department (GAD) has been highlighted at a cross government event.

Experts from GAD showcased our drought modelling work and the development of our pension dashboards to the Government Predictive Analytics Network (GPAN).

Wide network

GPAN was created in 2014 to share best practice in predictive analytics planning, strategy and innovation. It covers more than 15 government departments and GAD has contributed editorial support for the Predictive Analytics Handbook since 2014.

The Network’s objectives are to:

provide analytical leadership in the predictive analytics, data science, machine learning and artificial intelligence spaces

gather knowledge from across the Civil Service, private sector and academia to enable sharing with government analysts

promoting cross-sectoral collaboration

highlight best practice and promoting innovation

offer support to government wide data science initiatives

help promote the relevance and use of predictive analytics to operational colleagues in departments

Data scientists at GAD benefit from the insights of GPAN’s members. This includes using the Predictive Analytics Handbook that was developed by HMRC and members of the group. It provides a framework to help plan, specify, manage, build, deploy and assure predictive models.

Learn and contribute

Jon Day, who leads on data science in GAD said “The GPAN group provides an excellent opportunity to share perspectives and to learn from each other.

“We’ll continue to engage with the network to understand, learn and contribute to predictive analytical work in government and thereby help government deliver on its strategic ambitions.”

GAD continues to perform data science work and we are keen to partner with departments on projects and explore how data science can help them.

Press release distributed by Media Pigeon on behalf of GOV.UK, on Sep 26, 2022. For more information subscribe and follow