PyData brings together users and developers of data analysis tools in Python and other languages. At PyData Amsterdam, the community gathered to discuss how best to apply languages and tools to continuously evolving challenges in data management, processing, analytics, and visualization.
Just like in 2016, GoDataDriven headed the organizing comittee of PyData Amsterdam, bringing together over 300 Python enthusiasts at the Booking.com offices in Amsterdam.
Besides organizing, quite a few GoDataDriven consultant submitted proposals for presentations. The following presentations have been recorded and are now available online:
Rogier van der Geer talked about implementing Risk in Python and training a genetic algorithm to play the board game.
Niels Zeilemaker's presentation was all about deploying models to production. This can sometimes be more difficult than developing the model itself. That's why in this talk, Niels explains how we deploy our Pandas/Scikit machine learning models to production using Flask, Docker, and Kubernetes.
Next up, Giovanni Lanzani combined several talks that he gave at the PyData Amsterdam meetups. One of the main topics in this talk is the data science workflow and where things mostly go wrong.
The talk also touches on the type of people that can help companies do applied data science, and what most MOOCs don't teach you.
Interested in more PyData? Then make sure to join the PyData Amsterdam group on Meetup.