Making transport more robust and interpretable by moving data through a small number of anchor points
arXiv:2012.11589
Abstract
Optimal transport (OT) is a widely used technique for distribution alignment, with applications throughout the machine learning, graphics, and vision communities. Without any additional structural assumptions on trans-port, however, OT can be fragile to outliers or noise, especially in high dimensions. Here, we introduce a new form of structured OT that simultaneously learns low-dimensional structure in data while leveraging this structure to solve the alignment task. Compared with OT, the resulting transport plan has better structural interpretability, highlighting the connections between individual data points and local geometry, and is more robust to noise and sampling. We apply the method to synthetic as well as real datasets, where we show that our method can facilitate alignment in noisy settings and can be used to both correct and interpret domain shift.
References in corpus (12)
- Fast Computation of Wasserstein Barycenters
- A geometric analysis of subspace clustering with outliers
- Near-linear time approximation algorithms for optimal transport via Sinkhorn iteration
- Ground Metric Learning
- Hierarchical Optimal Transport for Document Representation
- Towards Optimal Transport with Global Invariances
- Multilevel Clustering via Wasserstein Means
- Statistical Optimal Transport via Factored Couplings
- Hierarchical Optimal Transport for Multimodal Distribution Alignment
- Learning Autoencoders with Relational Regularization
- Differentiable Deep Clustering with Cluster Size Constraints
- Fast and Robust Comparison of Probability Measures in Heterogeneous Spaces