Does smoking cause cancer?
Does AI identify infarctions better than a doctor?
Both these questions require an RCT (randomised control trial) in order to perform a true A|B test. In practice, however, RCTs are often expensive, time-consuming, need to be designed prospectively, and, most importantly, may be ethically impossible.
In this webinar, we will focus on Propensity score matching (PSM) - a fast and cheap statistical technique used on retrospective data, that approximates a true RCT. We will explore the assumptions, technique itself and implementation in a Python library called "psmpy".
Prerequisites: Statistics, Python basics and some data-sci libraries (numpy, pandas, matplotlib)
Date:
8.10.2026
Time:
17:00 -19:30
Language:
English
We will send an invitation for this webinar to your calendar, once you register :)
Lecturer:
Jakub Bahyl - Data Specialist at the Ministry of Health of Slovakia
LI:
https://www.linkedin.com/in/jakub-bahyl/