Causality for Machine Learning

Abstract

Graphical causal inference as pioneered by Judea Pearl arose from research on artificial intelligence (AI), and for a long time had little connection to the field of machine learning. This article discusses where links have been and should be established, introducing key concepts along the way. It argues that the hard open problems of machine learning and AI are intrinsically related to causality, and explains how the field is beginning to understand them.

References in corpus (56)

• Counterfactual Fairness
• Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations
• Kernel-based Conditional Independence Test and Application in Causal Discovery
• External Validity: From Do-Calculus to Transportability Across Populations
• Distinguishing cause from effect using observational data: methods and benchmarks
• Water vapour in the atmosphere of the habitable-zone eight Earth-mass planet K2-18 b
• On the Identifiability of the Post-Nonlinear Causal Model
• Towards the first adversarially robust neural network model on MNIST
• A Survey of Learning Causality with Data: Problems and Methods
• Stellar and Planetary Properties of K2 Campaign 1 Candidates and Validation of 17 Planets, Including a Planet Receiving Earth-like Insolation
• Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift
• Invariant Models for Causal Transfer Learning
• The Hardness of Conditional Independence Testing and the Generalised Covariance Measure
• On Causal and Anticausal Learning
• A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
• Inferring deterministic causal relations
• Representation Learning: A Review and New Perspectives
• Detecting and Correcting for Label Shift with Black Box Predictors
• Learning Independent Causal Mechanisms
• Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions
• Evaluating Reinforcement Learning Algorithms in Observational Health Settings
• Causal Reasoning from Meta-reinforcement Learning
• Recurrent Independent Mechanisms
• Identifiability of Causal Graphs using Functional Models
• Causal Consistency of Structural Equation Models
• Towards a Learning Theory of Cause-Effect Inference
• Identifying confounders using additive noise models
• Independently Controllable Features
• Telling cause from effect based on high-dimensional observations
• Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search
• Learning stable and predictive structures in kinetic systems: Benefits of a causal approach
• Deconfounding Reinforcement Learning in Observational Settings
• From Ordinary Differential Equations to Structural Causal Models: the deterministic case
• Domain Generalization via Conditional Invariant Representation
• Counterfactuals uncover the modular structure of deep generative models
• Optimal Decision Making Under Strategic Behavior
• Consistency of Causal Inference under the Additive Noise Model
• Telling cause from effect in deterministic linear dynamical systems
• Causal Generative Neural Networks
• Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness
• Group invariance principles for causal generative models
• Generalization in anti-causal learning
• Causal Markov condition for submodular information measures
• From Deterministic ODEs to Dynamic Structural Causal Models
• Algorithmic independence of initial condition and dynamical law in thermodynamics and causal inference
• Learning First-Order Symbolic Representations for Planning from the Structure of the State Space
• Semi-Supervised Learning, Causality and the Conditional Cluster Assumption
• Preventing Failures Due to Dataset Shift: Learning Predictive Models That Transport
• Competitive Training of Mixtures of Independent Deep Generative Models
• Detecting non-causal artifacts in multivariate linear regression models
• Fast Conditional Independence Test for Vector Variables with Large Sample Sizes
• Error Asymmetry in Causal and Anticausal Regression
• Causal and anti-causal learning in pattern recognition for neuroimaging
• The Arrow of Time in Multivariate Time Series
• Robust Learning via Cause-Effect Models
• Multi-Level Cause-Effect Systems

Cited by in corpus (69)

• Shortcut Learning in Deep Neural Networks
• Explainable Machine Learning for Scientific Insights and Discoveries
• Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems
• Interpretable Machine Learning -- A Brief History, State-of-the-Art and Challenges
• Underspecification Presents Challenges for Credibility in Modern Machine Learning
• Weakly-Supervised Disentanglement Without Compromises
• Deep Structural Causal Models for Tractable Counterfactual Inference
• Learning Invariant Representations for Reinforcement Learning without Reconstruction
• A causal view of compositional zero-shot recognition
• Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style
• On the Binding Problem in Artificial Neural Networks
• Learning Causal Semantic Representation for Out-of-Distribution Prediction
• A survey of algorithmic recourse: definitions, formulations, solutions, and prospects
• Socially Responsible AI Algorithms: Issues, Purposes, and Challenges
• Instance-dependent Label-noise Learning under a Structural Causal Model
• CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning
• Counterfactual Generative Networks
• Causal Navigation by Continuous-time Neural Networks
• Counterfactual Zero-Shot and Open-Set Visual Recognition
• White Paper Machine Learning in Certified Systems
• Distilling Causal Effect of Data in Class-Incremental Learning
• Disentangling User Interest and Conformity for Recommendation with Causal Embedding
• Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond
• Transparency, Auditability and eXplainability of Machine Learning Models in Credit Scoring
• Improving Generalizability of Fake News Detection Methods using Propensity Score Matching
• Urban Anomaly Analytics: Description, Detection, and Prediction
• Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game
• Towards Automating the AI Operations Lifecycle
• Algorithms for Causal Reasoning in Probability Trees
• Interventional Sum-Product Networks: Causal Inference with Tractable Probabilistic Models
• Size-Invariant Graph Representations for Graph Classification Extrapolations
• Latent Causal Invariant Model
• Causal Curiosity: RL Agents Discovering Self-supervised Experiments for Causal Representation Learning
• Proactive Pseudo-Intervention: Causally Informed Contrastive Learning For Interpretable Vision Models
• AdaRL: What, Where, and How to Adapt in Transfer Reinforcement Learning
• The Intriguing Relation Between Counterfactual Explanations and Adversarial Examples
• Causality-aware counterfactual confounding adjustment for feature representations learned by deep models
• Unsuitability of NOTEARS for Causal Graph Discovery
• Patterns, predictions, and actions: A story about machine learning
• Strategic Instrumental Variable Regression: Recovering Causal Relationships From Strategic Responses
• CausalCity: Complex Simulations with Agency for Causal Discovery and Reasoning
• Action-Sufficient State Representation Learning for Control with Structural Constraints
• Distinguishing Cause from Effect Using Quantiles: Bivariate Quantile Causal Discovery
• A polynomial-time algorithm for learning nonparametric causal graphs
• iReason: Multimodal Commonsense Reasoning using Videos and Natural Language with Interpretability
• Generative Interventions for Causal Learning
• Amortized learning of neural causal representations
• Embracing the Disharmony in Medical Imaging: A Simple and Effective Framework for Domain Adaptation
• Harmonization with Flow-based Causal Inference
• Causal Multi-Agent Reinforcement Learning: Review and Open Problems
• Optimization-based Causal Estimation from Heterogenous Environments
• Analysis of Drifting Features
• Learning Contextual Causality from Time-consecutive Images
• Causal World Models by Unsupervised Deconfounding of Physical Dynamics
• Identifying Causal Structure in Dynamical Systems
• Typing assumptions improve identification in causal discovery
• No Representation without Transformation
• Neural Networks for Learning Counterfactual G-Invariances from Single Environments
• Path-specific Effects Based on Information Accounts of Causality
• Finding Valid Adjustments under Non-ignorability with Minimal DAG Knowledge
• Learning Spatial Relationships between Samples of Patent Image Shapes
• Boxhead: A Dataset for Learning Hierarchical Representations
• Encoding Causal Macrovariables
• Towards intervention-centric causal reasoning in learning agents
• GalilAI: Out-of-Task Distribution Detection using Causal Active Experimentation for Safe Transfer RL
• Exploiting non-i.i.d. data towards more robust machine learning algorithms
• Physical System for Non Time Sequence Data
• Inferring Causal Direction from Observational Data: A Complexity Approach
• VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queries