May 10, 2017
MIT
Machine learning expert Jonas Peters of the University of Copenhagen presents “Four Lectures on Causality”.
Produced by the Laboratory for Information & Decision Systems (LIDS) of MIT (lids.mit.edu/) and Models, Inference & Algorithms of the Broad Institute (broadinstitute.org/mia).
Most of recent machine learning is focused on pure predictive performance, which has been a driving force behind its practical success. The question of causality (understanding why predictions work) has been somewhat left behind. This paradigm is incredibly important, because it can help understand things like which genes cause which diseases, and which policy affects which economic indicator, for example.
In the field of causality we want to understand how a system reacts under interventions (e.g. in gene knock-out experiments). These questions go beyond statistical dependences and can therefore not be answered by standard regression or classification techniques. In this tutorial you will learn about the interesting problem of causal inference and recent developments in the field. No prior knowledge about causality is required.
Part 2: We present three ideas that can be used to infer causal structure from data: (1) finding (conditional) independencies in the data, (2) restricting structural equation models and (3) exploiting the fact that causal models remain invariant in different environments.
To watch the rest of this presentation visit the following links:
Part 1: • Lectures on Causality:...
Part 2: • Lectures on Causality:...
Part 3: • Lectures on Causality:...
Part 4: • Lectures on Causality:...
Copyright Broad Institute, 2017. All rights reserved.
12 июл 2024
