Wasserstein barycenters are NP-hard to compute
arXiv:2101.01100 · doi:10.1137/21M1390062
Abstract
Computing Wasserstein barycenters (a.k.a. Optimal Transport barycenters) is a fundamental problem in geometry which has recently attracted considerable attention due to many applications in data science. While there exist polynomial-time algorithms in any fixed dimension, all known running times suffer exponentially in the dimension. It is an open question whether this exponential dependence is improvable to a polynomial dependence. This paper proves that unless P=NP, the answer is no. This uncovers a "curse of dimensionality" for Wasserstein barycenter computation which does not occur for Optimal Transport computation. Moreover, our hardness results for computing Wasserstein barycenters extend to approximate computation, to seemingly simple cases of the problem, and to averaging probability distributions in other Optimal Transport metrics.
to appear in SIAM Journal on Mathematics of Data Science (SIMODS)
References in corpus (6)
- Multilevel Clustering via Wasserstein Means
- Gradient descent algorithms for Bures-Wasserstein barycenters
- On the Complexity of Approximating Wasserstein Barycenter
- Hardness results for Multimarginal Optimal Transport problems
- Debiased Sinkhorn barycenters
- Estimating Barycenters of Measures in High Dimensions