Do you and your team have the coding and maths skills to help the NHS recover from the pandemic? If you believe you do then don’t miss out – you’ve got less than two months to enter our machine learning challenge, for your chance to see your model turned into a fully-functional app. 

The Health Economics Unit (HEU) launched the machine learning challenge in May 2021 on the Analyst X platform, in partnership with NHS England and NHS Improvement (NHSI) and NHSX.

The aim of the challenge is to create a predictive model that will help the NHS recover from the COVID-19 pandemic. Yes, this is a very open question. As well as coding and maths skills, our judging panel will be looking for an aptitude for formulating these questions and problems.

Healthcare commissioners often have a broad problem that needs solving, one that is not clearly specified in data science terms. It’s down to data scientists to make the question realistic from a data science perspective. That’s a skill we all need – turning a big problem into one that can be broken down into steps and addressed with data analysis.

Chris Mainey of NHS England and NHS Improvement, is one of the judges for the challenge. Chris said: “There are some key steps teams should take when building a question and considering when machine learning is appropriate to answer that question. This includes specifying the problem that needs to be solved, the data needed to address that problem and which method should be used.”

The challenge aims to encourage models that are explainable to the system. Challenge entries should be accompanied with evidence or reason(s) for all outputs and with explanations that are understandable to individual users.

The explanations should correctly reflect the system’s process for generating the output, and the system should only operate under conditions for which it was designed or when the system reaches a sufficient confidence in its output.

Any health and care team can enter the challenge, including local authorities, NHS staff and non-NHS teams. A team can be formed from more than one organisation and can also partner with academia.

Prizes on offer include:

• A write-up of the winning project by The Health Economics Unit for conference poster submission and publication on the AnalystX workspace as a blog
• A podcast with the winning team hosted by The Health Economics Unit for publication on the AnalystX workspace
• An app of the winning model to be created by SAS

Entrants will need to submit a PowerPoint or video presentation, accompanied by their coding work. We’re asking teams to submit their codes so we can share that knowledge with other areas in the NHS, to be replicated locally.

The judging panel will comprise:

• Ming Tang, NHS England and NHS Improvement
• Ben Goldacre, The DataLab, Oxford University
• Sarah Culkin, NHSX
• Chris Mainey, NHS England and NHS Improvement
• Mark Frankish, SAS
• And our very own Bruno Petrungaro

The criteria by which entries will be judged include the accuracy of the model, its potential to impact the behaviour of patients and healthcare professionals, and innovative use of real-world NHS data.

To find out more about the AnalystX Machine Learning Challenge, please click here.

To submit an entry please complete the ML challenge application form and return it to heu.analystx@nhs.net.