Selecting Data Augmentation for Simulating Interventions

Abstract

Machine learning models trained with purely observational data and the principle of empirical risk minimization \citep{vapnik_principles_1992} can fail to generalize to unseen domains. In this paper, we focus on the case where the problem arises through spurious correlation between the observed domains and the actual task labels. We find that many domain generalization methods do not explicitly take this spurious correlation into account. Instead, especially in more application-oriented research areas like medical imaging or robotics, data augmentation techniques that are based on heuristics are used to learn domain invariant features. To bridge the gap between theory and practice, we develop a causal perspective on the problem of domain generalization. We argue that causal concepts can be used to explain the success of data augmentation by describing how they can weaken the spurious correlation between the observed domains and the task labels. We demonstrate that data augmentation can serve as a tool for simulating interventional data. We use these theoretical insights to derive a simple algorithm that is able to select data augmentation techniques that will lead to better domain generalization.

References in corpus (19)

• mixup: Beyond Empirical Risk Minimization
• The Effectiveness of Data Augmentation in Image Classification using Deep Learning
• Deep Domain Confusion: Maximizing for Domain Invariance
• Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathology
• Domain Generalization via Invariant Feature Representation
• Group Equivariant Convolutional Networks
• Generalizing Across Domains via Cross-Gradient Training
• Invariant Risk Minimization
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• On Learning Invariant Representation for Domain Adaptation
• Invariant Models for Causal Transfer Learning
• Learning to Generalize: Meta-Learning for Domain Generalization
• Learning Robust Representations by Projecting Superficial Statistics Out
• Joint Causal Inference from Multiple Contexts
• DIVA: Domain Invariant Variational Autoencoders
• Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions
• Lagrangian Neural Networks
• Affinity and Diversity: Quantifying Mechanisms of Data Augmentation
• Support and Invertibility in Domain-Invariant Representations

Cited by in corpus (9)

• Out-of-Distribution Generalization via Risk Extrapolation (REx)
• Automatic Data Augmentation for Generalization in Deep Reinforcement Learning
• Learning Causal Semantic Representation for Out-of-Distribution Prediction
• Counterfactual Generative Networks
• Out-of-distribution Prediction with Invariant Risk Minimization: The Limitation and An Effective Fix
• Latent Causal Invariant Model
• Intervention Design for Effective Sim2Real Transfer
• Generative Interventions for Causal Learning
• Causality-inspired Single-source Domain Generalization for Medical Image Segmentation