Message Passing Algorithms for the Dirichlet Diffusion Tree

Explore this paper's citation graph

Summary

This work provides the first deterministic approximate inference methods for DDT models and shows excellent performance compared to the MCMC alternative on a density estimation problem, and significantly outperforms kernel density estimators.

Type
article
Published
2011-01-01
Cited by
4
References
25

Keywords

Computer science, Tree (set theory), Message passing, Latent Dirichlet allocation, Diffusion

References

Cited by

Related papers