Monthly Archives: September 2015

Video Club: Sackler Colloquium “Drawing Causal Inference from Big Data”

On Wednesdays 12:30-13:30 we will jointly watch some of the video recordings of presentations given at the recent Sackler Colloquium Drawing Causal Inference from Big Data of the National Academy of Sciences. (Lunch is not included – bring it yourself!)

Schedule: at the bottom of the Meetings page

Abstract: This colloquium was motivated by the exponentially growing amount of information collected about complex systems, colloquially referred to as “Big Data”. It was aimed at methods to draw causal inference from these large data sets, most of which are not derived from carefully controlled experiments. Although correlations among observations are vast in number and often easy to obtain, causality is much harder to assess and establish, partly because causality is a vague and poorly specified construct for complex systems. Speakers discussed both the conceptual framework required to establish causal inference and designs and computational methods that can allow causality to be inferred. The program illustrates state-of-the-art methods with approaches derived from such fields as statistics, graph theory, machine learning, philosophy, and computer science, and the talks will cover such domains as social networks, medicine, health, economics, business, internet data and usage, search engines, and genetics. The presentations also addressed the possibility of testing causality in large data settings, and will raise certain basic questions: Will access to massive data be a key to understanding the fundamental questions of basic and applied science? Or does the vast increase in data confound analysis, produce computational bottlenecks, and decrease the ability to draw valid causal inferences?

Talk by Elena Sokolova

You are all cordially invited to the next AMLab colloquium next Tuesday, September 29 at 16:00 in C3.163, where Elena Sokolova from the Intelligent Systems group at Radboud University will give a talk about “Causal discovery from medical data”. Afterwards we will have the usual drinks and snacks!

Title: Causal discovery from medical data
Speaker: Elena Sokolova
Abstract: The standard methods for data analysis of medical data involve statistical tests that tell whether the difference in means between two populations is statistically significant. These methods are easy to use, but are
restricted in the types of questions they can answer. Alternative approach to analyse medical data is causal discovery. It provides an opportunity to learn the causes and effects from data and it detects whether the dependency between variables is direct or mediated through other variables. In this talk I will present the challenges that one can face building a causal network from medical data and some interesting results that we inferred from this analysis that were not possible to obtain using standard statistical tests.

Video by Michael Jordan

Instead of a speaker, at this AMLab meeting we will watch a video of the recent Sackler Colloquium featuring Michael Jordan. Afterwards there will be drinks and snacks!

Date: Tuesday September 22, 2015
Time: 16:00-17:00
Location: SP C3.163

Video talk by: Michael Jordan, University of California: Berkeley

Title: On Computational Thinking, Inferential Thinking and Big Data

About the Sackler Colloquium: This colloquium was motivated by the exponentially growing amount of information collected about complex systems, colloquially referred to as “Big Data”. It was aimed at methods to draw causal inference from these large data sets, most of which are not derived from carefully controlled experiments. Although correlations among observations are vast in number and often easy to obtain, causality is much harder to assess and establish, partly because causality is a vague and poorly specified construct for complex systems. Speakers discussed both the conceptual framework required to establish causal inference and designs and computational methods that can allow causality to be inferred. The program illustrates state-of-the-art methods with approaches derived from such fields as statistics, graph theory, machine learning, philosophy, and computer science, and the talks will cover such domains as social networks, medicine, health, economics, business, internet data and usage, search engines, and genetics. The presentations also addressed the possibility of testing causality in large data settings, and will raise certain basic questions: Will access to massive data be a key to understanding the fundamental questions of basic and applied science? Or does the vast increase in data confound analysis, produce computational bottlenecks, and decrease the ability to draw valid causal inferences?