Detecting confounding in multivariate linear models via spectral analysis
Janzing, Dominik, Schoelkopf, Bernhard
We study a model where one target variable Y is correlated with a vector X:=(X_1,...,X_d) of predictor variables being potential causes of Y. We describe a method that infers to what extent the statistical dependences between X and Y are due to the influence of X on Y and to what extent due to a hidden common cause (confounder) of X and Y. The method relies on concentration of measure results for large dimensions d and an independence assumption stating that, in the absence of confounding, the vector of regression coefficients describing the influence of each X on Y typically has `generic orientation' relative to the eigenspaces of the covariance matrix of X. For the special case of a scalar confounder we show that confounding typically spoils this generic orientation in a characteristic way that can be used to quantitatively estimate the amount of confounding.
Apr-5-2017
- Country:
- North America > United States
- Indiana (0.04)
- Rhode Island > Providence County
- Providence (0.04)
- Oregon > Benton County
- Corvallis (0.04)
- New York > New York County
- New York City (0.04)
- Illinois > Cook County
- Chicago (0.04)
- California > San Diego County
- San Diego (0.04)
- Europe
- United Kingdom > England
- Cambridgeshire > Cambridge (0.04)
- Germany > Saarland
- Saarbrücken (0.04)
- United Kingdom > England
- Asia > Middle East
- Israel > Haifa District > Haifa (0.04)
- North America > United States
- Genre:
- Research Report (0.63)
- Industry:
- Materials (0.46)
- Technology: