Bayesian distance-to-set models replace latent variables with optimization-based projections to reduce dimensionality, improve MCMC efficiency, and establish posterior consistency plus an Occam's razor penalty for data near structured sets.
This perspective suggests that a range of popular machine learning procedures may be embedded into the divergence-to-set framework
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Bayesian Distance-to-Set Models: from Latent Variable to Latent Projection
Bayesian distance-to-set models replace latent variables with optimization-based projections to reduce dimensionality, improve MCMC efficiency, and establish posterior consistency plus an Occam's razor penalty for data near structured sets.