Recent Advances in Optimal Transport for Machine Learning (2023-06-28T00:00:00.000000Z)

TL;DR

This survey explores contributions of Optimal Transport for Machine Learning over the period 2012 – 2023, focusing on four sub-fields of Machine Learning: supervised, unsupervised, transfer and reinforcement learning.

Abstract

Recently, Optimal Transport has been proposed as a probabilistic framework in Machine Learning for comparing and manipulating probability distributions. This is rooted in its rich history and theory, and has offered new solutions to different problems in machine learning, such as generative modeling and transfer learning. In this survey we explore contributions of Optimal Transport for Machine Learning over the period 2012 – 2023, focusing on four sub-fields of Machine Learning: supervised, unsupervised, transfer and reinforcement learning. We further highlight the recent development in computational Optimal Transport and its extensions, such as partial, unbalanced, Gromov and Neural Optimal Transport, and its interplay with Machine Learning practice.

Authors

Eduardo Fernandes Montesuma

1 papers

Fred Ngolè Mboula

1 papers

A. Souloumiac

1 papers

TL;DR

Abstract

Authors

References203 items

Federated Wasserstein Distance

Federated Dataset Dictionary Learning for Multi-Source Domain Adaptation

Multi-Source Domain Adaptation through Dataset Dictionary Learning in Wasserstein Space

Spherical Sliced-Wasserstein

Revisiting Sliced Wasserstein on Images: From Vectorization to Convolution

Neural Optimal Transport

Optimal Transport Tools (OTT): A JAX Toolbox for all things Wasserstein

Don't Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergence

Foundations of Machine Learning

Improving Mini-batch Optimal Transport via Partial Transportation

An Invitation to Optimal Transport, Wasserstein Distances, and Gradient Flows

Wasserstein Barycenter Transport for Acoustic Adaptation

Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark

Testing Group Fairness via Optimal Transport Projections

Wasserstein Barycenter for Multi-Source Domain Adaptation

On the Applicability of Machine Learning Fairness Notions

OTCE: A Transferability Metric for Cross-Domain Cross-Task Representations

Unbalanced minibatch Optimal Transport; applications to Domain Adaptation

Wasserstein GANs Work Because They Fail (to Approximate the Wasserstein Distance)

Online Graph Dictionary Learning

Continuous Wasserstein-2 Barycenter Estimation without Minimax Optimization

A Riemannian Block Coordinate Descent Method for Computing the Projection Robust Wasserstein Distance

Dataset Dynamics via Gradient Flows in Probability Space

Fairness in Machine Learning: A Survey

Distributional Reinforcement Learning via Moment Matching

A Swiss Army Knife for Minimax Optimal Transport

Scalable Computations of Wasserstein Barycenter via Input Convex Neural Networks

Multi-source Domain Adaptation via Weighted Joint Distributions Optimal Transport

Projection Robust Wasserstein Distance and Riemannian Optimization

Severity-Aware Semantic Segmentation With Reinforced Wasserstein Training

OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport

Normalizing Flows: An Introduction and Review of Current Methods

A General Approach to Fairness with Optimal Transport

Importance-Aware Semantic Segmentation in Self-Driving with Discrete Wasserstein Training

Statistical and Topological Properties of Sliced Probability Divergences

Partial Optimal Transport with Applications on Positive-Unlabeled Learning

CO-Optimal Transport

Geometric Dataset Distances via Optimal Transport

How to train your neural ODE

A distributional code for value in dopamine-based reinforcement learning

Gromov-Wasserstein Factorization Models for Graph Clustering

Learning with minibatch Wasserstein : asymptotic and gradient properties

How Well Do WGANs Estimate the Wasserstein Metric?

Wasserstein-2 Generative Networks

Estimation of Wasserstein distances in the Spiked Transport Model

On the estimation of the Wasserstein distance in generative models

Optimal transport mapping via input convex neural networks

Advances in Domain Adaptation Theory

Wasserstein Distributionally Robust Optimization: Theory and Applications in Machine Learning

Distributionally Robust Optimization: A Review

Differentially Private Optimal Transport: Application to Domain Adaptation

Sinkhorn AutoEncoders

Asymptotic Guarantees for Learning Generative Models with the Sliced-Wasserstein Distance

Hierarchical Optimal Transport for Document Representation

Max-Sliced Wasserstein Distance and Its Use for GANs

Sliced Wasserstein Generative Models

Wasserstein Adversarial Regularization for Learning With Label Noise

Wasserstein Distance based Deep Adversarial Transfer Learning for Intelligent Fault Diagnosis

Generalized Sliced Wasserstein Distances

Subspace Robust Wasserstein distances

Interpolating between Optimal Transport and MMD using Sinkhorn Divergences

Sample Complexity of Sinkhorn Divergences

Dynamic Programming

Policy Optimization as Wasserstein Gradient Flows

Statistical Optimal Transport via Factored Couplings

Neural Ordinary Differential Equations

Obtaining Fairness using Optimal Transport Theory

Differential Properties of Sinkhorn Approximation for Learning with Wasserstein Distance

Optimal Transport for structured data with application on graphs

DeepJDOT: Deep Joint distribution optimal transport for unsupervised domain adaptation

Optimal Transport for Multi-source Domain Adaptation under Target Shift

Spectral Normalization for Generative Adversarial Networks

Optimal Transport: Fast Probabilistic Approximation with Exact Solvers

A Fast Proximal Point Method for Computing Exact Wasserstein Distance

Deep Visual Domain Adaptation: A Survey

Optimal Transport on Discrete Domains