Causal Inference via Kernel Deviance Measures
arXiv:1804.04622
Abstract
Discovering the causal structure among a set of variables is a fundamental problem in many areas of science. In this paper, we propose Kernel Conditional Deviance for Causal Inference (KCDC) a fully nonparametric causal discovery method based on purely observational data. From a novel interpretation of the notion of asymmetry between cause and effect, we derive a corresponding asymmetry measure using the framework of reproducing kernel Hilbert spaces. Based on this, we propose three decision rules for causal discovery. We demonstrate the wide applicability of our method across a range of diverse synthetic datasets. Furthermore, we test our method on real-world time series data and the real-world benchmark dataset Tubingen Cause-Effect Pairs where we outperform existing state-of-the-art methods.
References in corpus (3)
Cited by in corpus (15)
- Detecting crosstalk errors in quantum information processors
- Causal Confusion in Imitation Learning
- Learning Neural Causal Models from Unknown Interventions
- A Measure-Theoretic Approach to Kernel Conditional Mean Embeddings
- Higher Order Kernel Mean Embeddings to Capture Filtrations of Stochastic Processes
- Quantum Kernel Machine Learning With Continuous Variables
- Distinguishing Cause from Effect Using Quantiles: Bivariate Quantile Causal Discovery
- Leveraging directed causal discovery to detect latent common causes
- A polynomial-time algorithm for learning nonparametric causal graphs
- Formally Justifying MDL-based Inference of Cause and Effect
- A Bayesian Model for Bivariate Causal Inference
- Causal Mosaic: Cause-Effect Inference via Nonlinear ICA and Ensemble Method
- BayesIMP: Uncertainty Quantification for Causal Data Fusion
- Learning Data-adaptive Nonparametric Kernels
- Causal Discovery by Kernel Intrinsic Invariance Measure