Wavelet DPP kernels deliver improved continuous variance reduction and a discretization procedure that preserves decay rates for discrete ML subsampling tasks including rough objectives.
and Shorack, Galen R
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
New RSS-adapted L-estimators for quantiles, including a scalable rank-stratum component version, deliver efficiency gains over standard empirical estimators in simulations and a real NHANES application.
Extends BLASSO to multivariate GMMs with component-specific unknown diagonal covariances and derives non-asymptotic recovery guarantees under an explicit separation condition using Fisher-Rao geometry.
citing papers explorer
-
State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives
Wavelet DPP kernels deliver improved continuous variance reduction and a discretization procedure that preserves decay rates for discrete ML subsampling tasks including rough objectives.
-
L-Estimation of Population Quantiles Using Ranked Set Sampling
New RSS-adapted L-estimators for quantiles, including a scalable rank-stratum component version, deliver efficiency gains over standard empirical estimators in simulations and a real NHANES application.
-
Gaussian Mixture Model with unknown diagonal covariances via continuous sparse regularization
Extends BLASSO to multivariate GMMs with component-specific unknown diagonal covariances and derives non-asymptotic recovery guarantees under an explicit separation condition using Fisher-Rao geometry.