The robust recoverable information about x* from y under q-sparse adversarial e is the coset x* + ker(U) for U the projection onto the common rowspace of all A submatrices with 2q rows deleted; all minimizers of ||y - A x||_0 lie in this set, with a sharp exact-vs-trivial phase transition for i.i.d.
Low-rank pos- itive semidefinite matrix recovery from corrupted rank- one measurements
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IT 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
What Can Be Recovered Under Sparse Adversarial Corruption? Assumption-Free Theory for Linear Measurements
The robust recoverable information about x* from y under q-sparse adversarial e is the coset x* + ker(U) for U the projection onto the common rowspace of all A submatrices with 2q rows deleted; all minimizers of ||y - A x||_0 lie in this set, with a sharp exact-vs-trivial phase transition for i.i.d.