Difference of Convex Programming in the Wasserstein Space with Applications to MMD Optimization
cs.LG, math.OC
Submitted: 2026-06-26
Updated: 2026-11-09
Code: https://github.com/clbonet/Wasserstein_Convex_Concave_Procedure
Terminology
Sources
- Gradient Flow Drifting: Generative Modeling via Wasserstein Gradient Flows of KDE-Approximated Divergences
- A Lagrangian approach to totally dissipative evolutions in Wasserstein spaces
- Stochastic Difference-of-Convex Optimization with Momentum
- Quantitative Convergence of Wasserstein Gradient Flows of Kernel Mean Discrepancies
- Generative Modeling via Drifting
- Learning Monge maps with constrained drifting models
- A subdifferential characterization via Busemann functions and applications to DC optimization on Hadamard manifolds
- DC-LA: Difference-of-Convex Langevin Algorithm
- Continuous-Time Dynamics of the Difference-of-Convex Algorithm
- Forward-backward splitting under the light of generalized convexity
- Tight Analysis of Difference-of-Convex Algorithm (DCA) Improves Convergence Rates for Proximal Gradient Descent
- Accelerated gradient descent method for functionals of probability measures by new convexity and smoothness based on transport maps
- Convergence Rates for Distribution Matching with Sliced Optimal Transport
- Factored couplings in multi-marginal optimal transport via difference of convex programming
- Generative Drifting is Secretly Score Matching: a Spectral and Variational Perspective
- Forward-Euler time-discretization for Wasserstein gradient flows can be wrong
- Forward Euler for Wasserstein Gradient Flows: Breakdown and Regularization
- Random Coordinate Descent on the Wasserstein Space of Probability Measures
- Convergence Analysis of the Wasserstein Proximal Algorithm beyond Geodesic Convexity
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis