Rethinking the Diffusion Models for Numerical Tabular Data Imputation from the Perspective of Wasserstein Gradient Flow.
Zhichao ChenHaoxuan LiFangyikang WangOdin ZhangHu XuXiaoyu JiangZhihuan SongEric H. WangPublished in: CoRR (2024)
Keyphrases
- diffusion models
- data imputation
- gradient flow
- missing values
- diffusion model
- information diffusion
- energy functional
- missing data
- active contours
- active contour model
- level set
- social networks
- partial differential equations
- curve evolution
- variational formulation
- influence maximization
- multiresolution
- image dependent
- image processing
- machine learning
- social network analysis
- image analysis