Greedy Empirical Bayes for Density Estimation
Source: arxiv.org/abs/2611.22796
Bayesian nonparametric density estimation is challenging in high dimensions, where traditional approaches like the Dirichlet process (DP) mixture rely on computationally intensive Markov chain Monte Carlo or approximations. We introduce the partition posterior bootstrap (PPB), where a flexible prior is constructed by smoothing a DP over a partition of the sample space. Conditioned on this partition, exact posterior sampling reduces to a simple Bayesian bootstrap-like algorithm. Importantly, we select the partition using a greedy empirical Bayes (GEB) algorithm which scales to high dimensions…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2611.22796
curl -X POST /api/v1/records/arxiv/2611.22796/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2611.22796/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'