On the Identifiability of the Post-Nonlinear Causal Model

Abstract

By taking into account the nonlinear effect of the cause, the inner noise effect, and the measurement distortion effect in the observed variables, the post-nonlinear (PNL) causal model has demonstrated its excellent performance in distinguishing the cause from effect. However, its identifiability has not been properly addressed, and how to apply it in the case of more than two variables is also a problem. In this paper, we conduct a systematic investigation on its identifiability in the two-variable case. We show that this model is identifiable in most cases; by enumerating all possible situations in which the model is not identifiable, we provide sufficient conditions for its identifiability. Simulations are given to support the theoretical results. Moreover, in the case of more than two variables, we show that the whole causal structure can be found by applying the PNL causal model to each structure in the Markov equivalent class and testing if the disturbance is independent of the direct causes for each variable. In this way the exhaustive search over all possible causal structures is avoided.

Appears in Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence (UAI2009)

References in corpus (1)

• Gaussian Process Networks

Cited by in corpus (41)

• Kernel-based Conditional Independence Test and Application in Causal Discovery
• Inferring deterministic causal relations
• Identifiability of Causal Graphs using Functional Models
• Learning Causal Semantic Representation for Out-of-Distribution Prediction
• Telling cause from effect based on high-dimensional observations
• Sparse Linear Identifiable Multivariate Modeling
• DoWhy: Addressing Challenges in Expressing and Validating Causal Assumptions
• Estimation of Structural Causal Model via Sparsely Mixing Independent Component Analysis
• Sample-Efficient Reinforcement Learning via Counterfactual-Based Data Augmentation
• Detecting low-complexity unobserved causes
• A Local Method for Identifying Causal Relations under Markov Equivalence
• Causal Discovery with General Non-Linear Relationships Using Non-Linear ICA
• The Randomized Causation Coefficient
• Generalized Independent Noise Condition for Estimating Latent Variable Causal Graphs
• Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game
• Causal Discovery with Cascade Nonlinear Additive Noise Models
• Distinguishing Cause and Effect via Second Order Exponential Models
• Causal Inference from Slowly Varying Nonstationary Processes
• Latent Instrumental Variables as Priors in Causal Inference based on Independence of Cause and Mechanism
• Interventional Experiment Design for Causal Structure Learning
• Causal inference using deep neural networks
• Causal learning with sufficient statistics: an information bottleneck approach
• Learning latent causal graphs via mixture oracles
• GLACIAL: Granger and Learning-based Causality Analysis for Longitudinal Imaging Studies
• Quantifying intrinsic causal contributions via structure preserving interventions
• Conditionally-additive-noise Models for Structure Learning
• Learning Joint Nonlinear Effects from Single-variable Interventions in the Presence of Hidden Confounders
• On the Sample Complexity of Causal Discovery and the Value of Domain Expertise
• Causal Autoregressive Flows
• The Effect of Noise Level on Causal Identification with Additive Noise Models
• Estimation of Counterfactual Interventions under Uncertainties
• Efficient Bayesian network structure learning via local Markov boundary search
• Independent mechanism analysis, a new concept?
• On the Role of Entropy-based Loss for Learning Causal Structures with Continuous Optimization
• Causal Discovery from Conditionally Stationary Time Series
• On the Generative Utility of Cyclic Conditionals
• Estimation of Bivariate Structural Causal Models by Variational Gaussian Process Regression Under Likelihoods Parametrised by Normalising Flows
• Gradient-based Causal Structure Learning with Normalizing Flow
• Causality and Generalizability: Identifiability and Learning Methods
• Causal Identification with Additive Noise Models: Quantifying the Effect of Noise
• Parallel ensemble methods for causal direction inference