Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions

Abstract

An important goal common to domain adaptation and causal inference is to make accurate predictions when the distributions for the source (or training) domain(s) and target (or test) domain(s) differ. In many cases, these different distributions can be modeled as different contexts of a single underlying system, in which each distribution corresponds to a different perturbation of the system, or in causal terms, an intervention. We focus on a class of such causal domain adaptation problems, where data for one or more source domains are given, and the task is to predict the distribution of a certain target variable from measurements of other variables in one or more target domains. We propose an approach for solving these problems that exploits causal inference and does not rely on prior knowledge of the causal graph, the type of interventions or the intervention targets. We demonstrate our approach by evaluating a possible implementation on simulated and real world data.

arXiv commentCamera-ready version, to be published in the proceedings of Neural Information Processing Systems 2018 (NIPS*2018)

Cited by in corpus (48)

• A review of domain adaptation without target labels
• Underspecification Presents Challenges for Credibility in Modern Machine Learning
• Invariant Risk Minimization
• Interventional Few-Shot Learning
• A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
• Causality for Machine Learning
• Domain Generalization using Causal Matching
• Anchor regression: heterogeneous data meets causality
• Learning Causal Semantic Representation for Out-of-Distribution Prediction
• Towards a Theoretical Framework of Out-of-Distribution Generalization
• Few-shot Domain Adaptation by Causal Mechanism Transfer
• Invariant Risk Minimization Games
• When is invariance useful in an Out-of-Distribution Generalization problem ?
• Causality-based Feature Selection: Methods and Evaluations
• CASTLE: Regularization via Auxiliary Causal Graph Discovery
• Domain Adaptation as a Problem of Inference on Graphical Models
• Nonlinear Invariant Risk Minimization: A Causal Approach
• Selecting Data Augmentation for Simulating Interventions
• A Unifying Causal Framework for Analyzing Dataset Shift-stable Learning Algorithms
• Gradient-Based Neural DAG Learning
• Zero-Shot Recommender Systems
• Domain adaptation under structural causal models
• Semi-Supervised Learning, Causality and the Conditional Cluster Assumption
• Preventing Failures Due to Dataset Shift: Learning Predictive Models That Transport
• Invariant Representation Learning for Treatment Effect Estimation
• Out-of-Distribution Generalization Analysis via Influence Function
• Counterfactual Predictions under Runtime Confounding
• Latent Causal Invariant Model
• I-SPEC: An End-to-End Framework for Learning Transportable, Shift-Stable Models
• Explaining The Efficacy of Counterfactually Augmented Data
• Causality-aware counterfactual confounding adjustment for feature representations learned by deep models
• Accounting for Unobserved Confounding in Domain Generalization
• Risk Variance Penalization
• Linear Regression Games: Convergence Guarantees to Approximate Out-of-Distribution Solutions
• An Information-theoretic Approach to Distribution Shifts
• Improving Model Robustness Using Causal Knowledge
• The Connection between Out-of-Distribution Generalization and Privacy of ML Models
• Balance Regularized Neural Network Models for Causal Effect Estimation
• Independent mechanism analysis, a new concept?
• Stable predictions for health related anticausal prediction tasks affected by selection biases: the need to deconfound the test set features
• Learning Under Adversarial and Interventional Shifts
• Regularizing towards Causal Invariance: Linear Models with Proxies
• Towards Unbiased Visual Emotion Recognition via Causal Intervention
• Robust Learning in Heterogeneous Contexts
• Causal Transfer Random Forest: Combining Logged Data and Randomized Experiments for Robust Prediction
• Causality and Generalizability: Identifiability and Learning Methods
• Adaptive Multi-Source Causal Inference
• Sample Observed Effects: Enumeration, Randomization and Generalization