An Empirical Study of Invariant Risk Minimization
arXiv:2004.05007
Abstract
Invariant risk minimization (IRM) (Arjovsky et al., 2019) is a recently proposed framework designed for learning predictors that are invariant to spurious correlations across different training environments. Yet, despite its theoretical justifications, IRM has not been extensively tested across various settings. In an attempt to gain a better understanding of the framework, we empirically investigate several research questions using IRMv1, which is the first practical algorithm proposed to approximately solve IRM. By extending the ColoredMNIST experiment in different ways, we find that IRMv1 (i) performs better as the spurious correlation varies more widely between training environments, (ii) learns an approximately invariant predictor when the underlying relationship is approximately invariant, and (iii) can be extended to an analogous setting for text classification.
arXiv commentPresented at the ICML 2020 Workshop on Uncertainty and Robustness in Deep Learning. Code at https://github.com/kakaobrain/irm-empirical-study
References in corpus (3)
Cited by in corpus (7)
- Unshuffling Data for Improved Generalization
- Fairness and Robustness in Invariant Learning: A Case Study in Toxicity Classification
- Meta-Learned Invariant Risk Minimization
- Predict then Interpolate: A Simple Algorithm to Learn Stable Classifiers
- Detecting Spurious Correlations with Sanity Tests for Artificial Intelligence Guided Radiology Systems
- Causal Learning for Socially Responsible AI
- An Empirical Framework for Domain Generalization in Clinical Settings