Dirichlet draws are sparse with high probability
arXiv:1301.4917v112 citations
Originality Synthesis-oriented
AI Analysis
This addresses a theoretical gap in probability and statistics, but it is incremental as it formalizes a known folklore result.
The paper tackles the problem of understanding sparsity in draws from Dirichlet distributions with parameters less than 1, providing an elementary proof that such draws are typically sparse, with most coordinates being small.
This note provides an elementary proof of the folklore fact that draws from a Dirichlet distribution (with parameters less than 1) are typically sparse (most coordinates are small).