CountsDiff is a diffusion model designed to operate on natural numbers for generating and imputing count-based data. The paper extends diffusion model capabilities to discrete count data, a previously underexplored area. Applications include statistical modeling, survey analysis, and any domain with discrete count variables.
Research
CountsDiff: A Diffusion Model on the Natural Numbers for Generation and Imputation of Count-Based Data
Diffusion models now generate and impute discrete count data, bridging generative AI techniques from continuous domains into statistical applications like survey analysis and missing value imputation.
Tuesday, April 7, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.LG (Machine Learning)BY sys://pipeline
Tags
research