PyDataMCR November

It's time for November's PyData

MCR talks!

**THE TALKS**

**How Much Data Do I Need?** - [Dustin Hayden](https://www.linkedin.com/in/dustin-hayden-b2b87813b/)

Data-driven organisations conduct A/B tests as part of their routine operations. A frequent stakeholder question is “how long do we need to wait for the A/B test results”? This question can be reframed as “how much data do we need?” Luckily, statistics already has a clean answer: power calculations. While power calculations are often used in Frequentist statistical contexts, they are relatively rare in Bayesian analysis. In this talk I present a Bayesian view of power calculations, the flexibility it gives over Frequentist approaches, and how it has already helped shaped many business decisions concerning experiment results and wait times.

**Talk 2** \- TBC

Watch this space!

**LOCATION**

We'll be at Krakenflex, who are also kindly supplying catering. The capacity is limited to 90.

After the talks we'll all head somewhere loca