Abstract
Structure learning algorithms aim to retrieve the true causal structure from a set of observations. Most times only an equivalence class can be recovered and a unique model cannot be singled out. We hypothesized that casual directions could be inferred from the assessment of the strength of potential causal effects and such assessment can be computed by intervals comparison strategies. We introduce SLICE (Structural Learning with Intervals of Causal Effects), a new algorithm to decide on unresolved relations, which taps on the computation of causal effects and an acceptability index; a strategy for intervals comparison. For validation purposes, synthetic datasets were generated varying the graph size and density with samples drawn from Gaussian and non-Gaussian distributions. Comparison against LiNGAM is made to establish the performance of SLICE over 1440 scenarios using the normalised structural Hamming distance (SHD). The retrieved structures with SLICE showed smaller SHD values in the Gaussian case, improving the structure of the retrieved causal model in terms of correctly found directions. The acceptability index is a good predictor of the true causal effects (R2 = 0.62). The proposed strategy represents a new tool for discovering unravelled causal relations in the presence of observational data only.
Original language | English |
---|---|
Pages (from-to) | 296-307 |
Number of pages | 12 |
Journal | Proceedings of Machine Learning Research |
Volume | 72 |
Publication status | Published - 2018 |
Event | 9th International Conference on Probabilistic Graphical Models, PGM 2018 - Prague, Czech Republic Duration: 11 Sept 2018 → 14 Sept 2018 |
Bibliographical note
Funding Information:This research has been funded by the Mexican research council through project CB-2014-01-237251 and SMH is supported by a PhD scholarship from the CONACYT.
Publisher Copyright:
© 2018 Proceedings of Machine Learning Research. All rights reserved.
Keywords
- causal discovery
- causal effects
- structure learning
ASJC Scopus subject areas
- Artificial Intelligence
- Software
- Control and Systems Engineering
- Statistics and Probability