PyData @ Heineken - Beer quality w. spectral data & Practical forecasting

Is Python your second language? If so, then we have a very exciting announcement for you! Join us on the 19th of October, from 17.30 till 22.00, at the Heineken Experience to enjoy inspiring talks and network with fellow Python enthusiasts.

During this event, we will host two talks: one providing a practical (but not just basic!) overview of forecasting methods, the other on predicting beer quality using chemical analysis data. You don’t want to miss those! Before and after the presentations, there will be an opportunity for networking while enjoying a beer and snacks.

Schedule:
17.30-18.30: Welcome with food and drinks
18.30-19.15: Talk 1 - Forecasting: a practical overview of methods, Tomislav Suhina
19.15-19.30: Break
19.30-20.15: Talk 2 – Inferring beer quality from spectral chemical data analysis, Jurgen Nijkamp
20.15-22.00: Networking

“Forecasting: a practical overview of methods” by Tomislav Suhina
Opportunities to apply time-series forecasting are everywhere. Some examples could be planning production to meet future demand, predicting the number of visitors to make informed staffing decisions, or projecting any decision/curiosity-related quantity/kpi into the future.
There exists a truly sizeable list of algorithms implemented through various open-source packages at our disposal, ready to help us in various forecasting tasks, each with their own strengths, weaknesses, and compromises. Decision on when to use what is not always simple.
In this talk, I'll use toy datasets to share some of my learnings on different (simulated) forecasting scenarios, with each scenario having its own set of challenges and limitations.

“Inferring beer quality from spectral chemical data analysis” by Jurgen Nijkamp
Heineken has a long history of quality control for beer. Our flagship beer is regularly tested and compared across breweries to ensure it tastes the same, no matter where you buy it. The quality is evaluated using both chemical analysis and via a panel of trained testers. This talk highlights the use of machine learning on spectral chemical analysis data to infer beer quality and the drivers that make beer age well.