PyData Eindhoven Edition

Wat we doen

Since we’re so excited for Pydata, we’re organizing this meetup to get warmed up in the sunny southern city of light Eindhoven!

--- Program for the meetup
18:00 Doors open & Dinner
18:40 Introduction
18:45 Talk 1) Tim Butterbrod - Getting Started with Active Learning Using fashion MNIST
19:10 Break
19:15 Talk 2) Stijn Janssen - Learnings from our experiences with Airflow
19:45 Break
19:50 Talk 3) Vincent Warmerdam - Open-source Contributions
20:15 Drinks and networking

While neural networks thrive on an abundance of labeled data, in reality such an abundance of data is rarely available. And labeling data, whether in an online setting like on a website, or in fact letting data be labeled by humans is costly and time consuming. Active learning studies how we can use algorithms to strategically choose which data points to label. Tim Butterbrod will show you how to get started with active learning using the fashion mnist data set!

In the second talk, Stijn Janssen will share from experiences of moving our (EMR) job scheduling to Airflow. He will shares learning that hopefully will be interesting both when considering starting to use Airflow for job scheduling, or for the continuous improvement process. From our experience, there are several topics you want to carefully decide about based on your specific situation!

In the third talk, Vincent Warmerdam will share about his opensource contribution, which include scikit-lego (https://scikit-lego.readthedocs.io/en/latest/index.html) for customizing sklearn Pipelines and brent (https://github.com/koaning/brent) for exploring graphical causal models and do-calculus.