ZIN: When and How to Learn Invariance Without Environment Partition?
This work addresses the challenge of robust model learning in heterogeneous data settings for machine learning practitioners, but it is incremental as it builds on existing invariant learning methods by adding partition learning.
The paper tackles the problem of learning invariant features from heterogeneous data without predefined environment partitions, showing it's fundamentally impossible without inductive biases or additional information, and proposes a framework that jointly learns partitions and invariant representations, achieving improved performance validated on synthetic and real-world datasets.
It is commonplace to encounter heterogeneous data, of which some aspects of the data distribution may vary but the underlying causal mechanisms remain constant. When data are divided into distinct environments according to the heterogeneity, recent invariant learning methods have proposed to learn robust and invariant models based on this environment partition. It is hence tempting to utilize the inherent heterogeneity even when environment partition is not provided. Unfortunately, in this work, we show that learning invariant features under this circumstance is fundamentally impossible without further inductive biases or additional information. Then, we propose a framework to jointly learn environment partition and invariant representation, assisted by additional auxiliary information. We derive sufficient and necessary conditions for our framework to provably identify invariant features under a fairly general setting. Experimental results on both synthetic and real world datasets validate our analysis and demonstrate an improved performance of the proposed framework over existing methods. Finally, our results also raise the need of making the role of inductive biases more explicit in future works, when considering learning invariant models without environment partition. Codes are available at https://github.com/linyongver/ZIN_official .