Name: Challenging Google's PPC bidding algorithm using PyMC
Start: 2024-06-26T18:00:00+01:00
End: 2024-06-26T20:00:00+01:00
Location: Bayes Business School

**Abstract**: Pay per click (PPC) is a model of digital advertising where the advertiser pays a fee whenever their ad is clicked. Optimising how much to bid on a given keyword is fundamental to improving traffic to a company's website and to remain competitive. This talk provides an overview of a project where we implemented a dynamic bidding technique using a combination of Thompson Sampling using PyMC and an optimiser. We also discuss the ethical considerations of this project.

**Bios & Names (in alphabetical order)**

Dr Jordan Palmer is a Data Scientist at Datasparq. His academic research background focused on experimental dark matter physics and statistical discovery techniques. He has been recently focusing on pricing solutions using reinforcement learning techniques at Datasparq.

Dr Tom Prescott is a Data Scientist at Datasparq. His academic research background focused on applied mathematics, control engineering and computational statistics. Outside of academia, he has previously worked in government OR, agent-based simulations in digital twins, and revenue management.

Markus Gesmann

Bayesian Mixer London

Technology

Statistical Computing

Bayesian Statistics

Applied Statistics

Statistical Science

Statistical Modeling

Challenging Google's PPC bidding algorithm using PyMC

Bayes Business School

Share

Bayesian Mixer London

Challenging Google's PPC bidding algorithm using PyMC

Bayesian Mixer London

Details

Members are also interested in