this is for holding javascript data
Richard Guo edited untitled.tex
over 8 years ago
Commit id: 8f10dd2806aef18e951b79e766aa21584e9cee25
deletions | additions
diff --git a/untitled.tex b/untitled.tex
index 4d5cfb9..540f85d 100644
--- a/untitled.tex
+++ b/untitled.tex
...
\begin{enumerate}
\item The dictionary $\mathbf{A} \in \mathbb{R}^{n \times n}$ is a square and invertible.
\item The elements in $\mathbf{X}_0$ is independently drawn from Bernoulli-Gaussian (BG) model. $X_{0{[ij]}} = \Omega_{ij} V_{ij}$, where $\Omega_{ij} \sim \text{Ber}(\theta)$ and $V_{ij} \sim N(0,1)$.
\end{enumerate}
The main result is that when $\theta \in (0, 1/3)$, $\mathbf{A}_0$ and $\mathbf{X}_0$ can be exactly recovered with a polynomial time algorithm with high probability, if number of samples $p$ is no less than a polynomial of $(n, 1/\theta, \kappa(\mathbf{A}_0), 1/\mu)$, where $\kappa(\mathbf{A}_0)$ is the conditional number of the dictionary.
\end{enumerate}