Dirichlet draws are sparse with high probability

Abstract

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).

4 pages

Cited by in corpus (5)

• Improving the Gaussian Mechanism for Differential Privacy: Analytical Calibration and Optimal Denoising
• A Tensor Approach to Learning Mixed Membership Community Models
• Contrastive learning, multi-view redundancy, and linear models
• Improving application performance with biased distributions of quantum states
• On some provably correct cases of variational inference for topic models