New mathematical distances for partial optimal transport unveiled

Researchers have introduced new mathematical tools, entropic partial optimal transport and a partial mixture Gromov--Wasserstein distance, designed to compare probability measures and metric measure spaces. These methods address limitations of existing balanced formulations by allowing for partial matching, which is beneficial for datasets with outliers or incomplete information. The paper details the mathematical properties of these new distances, including their existence, uniqueness, and limits, and demonstrates their application through numerical experiments on synthetic data and point clouds. AI

RANK_REASON The cluster contains a new academic paper detailing novel mathematical methods. [lever_c_demoted from research: ic=1 ai=0.4]

Read on arXiv stat.ML →

• Entropic partial optimal transport
• Gaussian components
• Gaussian Mixture Models
• Gromov--Wasserstein distances
• optimal transport
• Partial Gromov--Wasserstein distance
• point cloud
• Toshiaki Yachimura

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →