Doubly sparse spiked Wigner model
A generalization of the sparse spiked Wigner model in which both the signal and the noise have sparse structure; its theoretical conditions for distinguishability, support recovery and strong recovery are studied.
Timeline (last 90 days)
A recent study systematically characterized the distinguishability and recovery conditions in the doubly sparse setting, covering almost the entire parameter space.
The core setup of the study is that the signal and the noise each have independent sparsity parameters while the signal-to-noise ratio remains of constant order—the signal strength does not grow as the matrix size increases.
The doubly sparse model simultaneously parametrizes the nonzero support of the signal and the nonzero structure of the noise, examining how the BBP-type phase transition moves with the two types of sparsity.
The paper explicitly positions itself as the last piece of the 'jigsaw puzzle', and together with the authors' earlier results on principal component analysis (PCA) (Dumitriu et al., 2026), the current analysis covers almost all parameters and mechanisms of distinction and recovery in the doubly sparse spiked Wigner model.
All Facts (4)
A recent study systematically characterized the distinguishability and recovery conditions in the doubly sparse setting, covering almost the entire parameter space.
50%UnverifiedThe core setup of the study is that the signal and the noise each have independent sparsity parameters while the signal-to-noise ratio remains of constant order—the signal strength does not grow as the matrix size increases.
50%UnverifiedThe doubly sparse model simultaneously parametrizes the nonzero support of the signal and the nonzero structure of the noise, examining how the BBP-type phase transition moves with the two types of sparsity.
50%UnverifiedThe paper explicitly positions itself as the last piece of the 'jigsaw puzzle', and together with the authors' earlier results on principal component analysis (PCA) (Dumitriu et al., 2026), the current analysis covers almost all parameters and mechanisms of distinction and recovery in the doubly sparse spiked Wigner model.
50%