In the sublinear sparsity limit the ML estimator achieves vanishing squared error below a noise threshold that coincides with the converse bound for constant-amplitude signals, proving asymptotic optimality of separable Bayesian estimators.
Th e mutual information in random linear estimation beyond i.i.d. matr ices,
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.IT 2verdicts
UNVERDICTED 2representative citing papers
GOAMP achieves error-free reconstruction of sublinearly sparse signals from linear measurements when the measurement dimension exceeds a threshold matching that of Gaussian AMP, provided the non-zero support avoids a neighborhood of the origin.
citing papers explorer
-
Direct and Converse Theorems in Estimating Signals with Sublinear Sparsity
In the sublinear sparsity limit the ML estimator achieves vanishing squared error below a noise threshold that coincides with the converse bound for constant-amplitude signals, proving asymptotic optimality of separable Bayesian estimators.
-
Generalized Orthogonal Approximate Message-Passing for Sublinear Sparsity
GOAMP achieves error-free reconstruction of sublinearly sparse signals from linear measurements when the measurement dimension exceeds a threshold matching that of Gaussian AMP, provided the non-zero support avoids a neighborhood of the origin.