A novel (pre-)metric for causal graphs

Add to your list(s) Download to your calendar using vCal

• Jonas Peters, ETH Zürich
• Friday 30 January 2015, 14:30-15:30
• MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.

If you have a question about this talk, please contact .

Causal discovery methods estimate a system’s underlying causal graph from observational data. But even if the true structure is known, it is not clear how to evaluate the performance of this estimator. In many cases, we use the causal graph to infer the system’s behaviour under specific interventions, i.e. manipulations of some of the variables. Different graphs may then result in different predicted behaviours. The structural intervention distance (SID) quantifies these differences and thereby defines a (pre-)metric between graphs. We briefly discuss recent results on identification of causal structure and show that they are closely related to the concept of SID . This talk does not require any prior knowledge about causal concepts.

This talk is part of the Statistics series.

This talk is included in these lists:

• All CMS events
• All Talks (aka the CURE list)
• CMS Events
• Cambridge Forum of Science and Humanities
• Cambridge Language Sciences
• Cambridge talks
• Chris Davis' list
• DPMMS Lists
• DPMMS info aggregator
• DPMMS lists
• Guy Emerson's list
• Hanchen DaDaDash
• Interested Talks
• MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge
• Machine Learning
• School of Physical Sciences
• Statistical Laboratory info aggregator
• Statistics
• Statistics Group
• bld31
• custom
• rp587

Note that ex-directory lists are not shown.