Welcome to the first Bayesian Mixer of 2025! 

John Whitamore will present on Bayesian inference and learning on a graph: use cases, theory and implementation

Abstract: How can a retail demand forecasting system share insights between the same product sold in many different stores? Or between similar products within one store? And how can that demand forecasting system share insights with related models, used for ordering, inventory levels, pricing and so on?

Similarly, how can a financial trading system share insights between securities and derivatives which move together? How can risk management systems share insights about risk exposures?

This talk discusses use cases, theory and implementation of Bayesian inference and learning on a graph - an approach with applicability to solving practical problems across a wide range of industries.

Bio: John Whitamore is a data scientist leading a startup currently in stealth mode. He was previously Head of Data Science for the food business of Marks and Spencer. He has also served as Head of Trading and Systems for an investment management company and as Global Head of Convertible Bond Trading for a European investment bank.

Thanks: Many thanks to Bayes Business School for hosting the Bayesian Mixer!