Tengyuan Liang

Computational and statistical boundaries for submatrix localization in a large noisy matrix

The interplay between computational efficiency and statistical accuracy in high-dimensional inference has drawn increasing attention in the literature. In this paper, we study computational and statistical boundaries for submatrix localization. Given one observation of (one or multiple non-overlapping) signal submatrix (of magnitude

Law of log determinant of sample covariance matrix and optimal estimation of differential entropy for high-dimensional Gaussian distributions

\lambda

Geometric Inference for General High-Dimensional Linear Inverse Problems

and size

Escaping the Local Minima via Simulated Annealing: Optimization of Approximately Convex Functions

k_m \times k_n

Fisher-Rao Metric, Geometry, and Complexity of Neural Networks.

) contaminated with a noise matrix (of size

Learning with Square Loss: Localization through Offset Rademacher Complexity

m \times n

Interaction Matters: A Note on Non-asymptotic Local Convergence of Generative Adversarial Networks.

), we establish two transition thresholds for the signal to noise

Inference via Message Passing on Partially Labeled Stochastic Block Models

\lambda/\sigma

On Zeroth-Order Stochastic Convex Optimization via Random Walks.

ratio in terms of

Geometrizing Local Rates of Convergence for Linear Inverse Problems

m