| ID:
596824.0,
MPI für Intelligente Systeme (ehemals Max-Planck-Institut für Metallforschung) / Abt. Schölkopf (Empirical Inference) |
| Testing whether linear equations are causal: A free probability theory approach |
| Authors: | Zscheischler, J.; Janzing, D.; Zhang, K. |
| Publisher: | AUAI Press |
| Place of Publication: | Corvallis, OR, USA |
| Date of Publication (YYYY-MM-DD): | 2011-07-01 |
| Title of Proceedings: | 27th Conference on Uncertainty in Artificial Intelligence (UAI 2011) |
| Start Page: | 839 |
| End Page: | 847 |
| Physical Description: | 8 |
| Place of Conference/Meeting: | Barcelona, Spain |
| Review Status: | not specified |
| Audience: | Not Specified |
| Intended Educational Use: | No |
| Abstract / Description: | We propose a method that infers whether linear relations between two high-dimensional variables X and Y are due to a causal influence from X to Y or from Y to X. The earlier proposed so-called Trace Method is extended to the regime where the dimension of the observed variables exceeds the sample size. Based on previous work, we postulate
conditions that characterize a causal relation between X and Y . Moreover, we describe a statistical test and argue that both causal directions are typically rejected if there is a common cause. A full theoretical analysis is
presented for the deterministic case but our approach seems to be valid for the noisy case, too, for which we additionally present an approach based on a sparsity constraint. The discussed method yields promising results for both simulated and real world data. |
| External Publication Status: | published |
| Document Type: | Conference-Paper |
| Communicated by: | Heide Klooz |
| Affiliations: | MPI für Intelligente Systeme/Abt. Schölkopf
|
| Identifiers: | URL:http://www.kyb.tuebingen.mpg.de/fileadmin/user_upl...
LOCALID:ZscheischlerJZ2011 |
|
Advanced
