AI Paper Atlas

Research map / Learning theory and algorithms / Optimal transport

Functional optimal transport: regularized map estimation and domain adaptation for functional data

Jiacheng Zhu, Aritra Guha, Dat Do, Mengdi Xu, XuanLong Nguyen, Ding Zhao

JMLR 2024 · Learning theory and algorithms › Optimal transport · approach: Transport map estimation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use FOT instead of GPOT, LSOT and DSOT for map estimation on mixture-of-sinusoid curvesFunctional optimal transport · Synthetic mixture of sinusoidal functionsWasserstein-based objective LImproved
Treat curves as functions rather than point-cloud vectors (FOT vs LSOT and DSOT)FOT · Synthetic sinusoids with non-continuous noisesmoothness of pushforwardImproved
Replace Gaussian-process OT assumption with model-free FOTFOT vs GPOT · Synthetic bimodal targetmultimodality captureImproved
Add power regularization instead of entropic regularization on the coupling (ablation)FOT · Synthetic sinusoidscoupling sparsityImproved
Estimate basis functions with FPCA instead of randomly misspecified basis (ablation)FOT · Synthetic Hermite-basis datamap qualityImproved
Increase number of basis eigenfunctions K beyond true K* of 15 (ablation)FOT · Synthetic map with K1*=K2*=15learning objectiveNo clear effect
Increase sample size n (ablation)FOT · Synthetic sinusoids, K=30, 10 repeatsFrobenius norm to optimal mapImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy