Research Seminar: "A first look at causal inference"

Program

Abstract:

One of the most vital steps in data analysis involves determining what has an effect on what. While the most effective way to answer this question is often through experimentation (eg randomized controlled trials and A/B testing), such methods can prove costly, unethical or unfeasible. Consequently, we tend to work with observational data that have not been gathered through testing. However, analyses of this nature can lead us, even in the simplest of cases, down erroneous paths or culs-de-sac. Causal inference is a working framework and a set of tools that provide us with model transparency, risk awareness and mathematical equations for removing bias. During this talk, we will take a look at some examples of how, through causal modelling, we draw conclusions that differ greatly from the results we obtain when causality is overlooked. We will also cover the contexts in which causal inference is necessary, the risks it entails, and the languages and tools that make use of it.

Bio: 

Aleix Ruiz de Villa earned his PhD in Mathematics from the Universitat Autònoma de Barcelona (UAB) and has since headed the data science departments at LaVanguardia.com, SCRM (the digital group responsible for Lidl’s mobile app) and Onna. He came up with the idea for the Barcelona Data Science and Machine Learning Meetup in 2014, co-founded the Barcelona R users group, in which he was involved from 2011 to 2017, and is on the Board of the Catalan Society of Mathematics.

Language: English