Jonas Peters
Wednesday 28th January 2015
Time: 4pm
B10 Basement Floor Seminar Room
Alexandra House, 17 Queen Square, London, WC1N 3AR
Causal Inference using Invariant Prediction
Why are we interested in the causal structure of a data-generating
process? In a classical regression problem, for example, we include a
variable into the model if it improves the prediction; it seems that no
causal knowledge is required. In many situations, however, we are
interested in the system's behavior under a change of environment, i.e.,
under a different distribution. Here, causal structures become important
because they are usually considered invariant under those changes and
can therefore be used to answer questions about this new distribution.
For example, a causal prediction (which uses only direct causes of the
target variable as predictors) remains valid even if we intervene on
predictor variables or change the whole experimental setting. We propose
a method that exploits the invariance principle when data from different
environments are available. This talk concentrates on ideas and concepts
and does not require any prior knowledge about causality.