Are basic QI tools lost in the pursuit of AI?

Date: Tuesday 12 December 2023, 3.00PM
Location: Royal Statistical Society, London, Errol Street
Section Group Meeting


Share this event

Omar McCarthy who will present the following:-
Are you really doing enough to optimise productivity in the increasing age of Data, Digital and Technology?
Hear an example of how Data Science/Machine Learning is being used to drive Operational Excellence within a Quality organisation.

Blaise Egan who will present the following:-
The recent fuss (hype?) around AI as a ground-breaking tool for revolutionising many aspects of business has emphasised the capabilities of modern machine learning (ML) techniques, but has taken the spotlight off more traditional statistical methods, which are not part of the training of most data scientists. Statisticians can have a major role to play in bringing such techniques to the fore, where appropriate.

Some examples from experience will be given:
  • Use of (highly) nonlinear models where linear models are suitable
  • Failure to use control groups in a designed experiment
  • Failure to realise that a comparison of two classifiers on the same dataset is a paired comparison
  • Regularly-scheduled re-estimation of a churn model where control charts would be more appropriate
  • Use of CUSUMs for Key Performance Indicator trends

Jonathan Smyth-Renshaw who will present the following:-
Is the fundamental principle of any process to achieve a defined target with minimum variation, aided or hindered by AI?

What are the boundary conditions between using AI and other techniques without the 'complexity of AI'?

Chair: Shirley Coleman

 
Fellow: free to attend

Non Fellow: £10