Fast global convergence of gradient methods for high-dimensional statistical recovery
arXiv:1104.4824
Abstract
Many statistical -estimators are based on convex optimization problems formed by the combination of a data-dependent loss function with a norm-based regularizer. We analyze the convergence rates of projected gradient and composite gradient methods for solving such problems, working within a high-dimensional framework that allows the data dimension $\pdim$ to grow with (and possibly exceed) the sample size $\numobs$. This high-dimensional structure precludes the usual global assumptions---namely, strong convexity and smoothness conditions---that underlie much of classical optimization analysis. We define appropriately restricted versions of these conditions, and show that they are satisfied with high probability for various statistical models. Under these conditions, our theory guarantees that projected gradient descent has a globally geometric rate of convergence up to the \emph{statistical precision} of the model, meaning the typical distance between the true unknown parameter and an optimal solution . This result is substantially sharper than previous convergence results, which yielded sublinear convergence, or linear convergence only up to the noise level. Our analysis applies to a wide range of -estimators and statistical models, including sparse linear regression using Lasso (-regularized regression); group Lasso for block sparsity; log-linear models with regularization; low-rank matrix recovery using nuclear norm regularization; and matrix decomposition. Overall, our analysis reveals interesting connections between statistical precision and computational efficiency in high-dimensional estimation.
References in corpus (10)
- Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization
- Simultaneous analysis of Lasso and Dantzig selector
- A Simpler Approach to Matrix Completion
- On the conditions used to prove oracle results for the Lasso
- Restricted strong convexity and weighted matrix completion: Optimal bounds with noise
- Estimation of high-dimensional low-rank matrices
- Robust Principal Component Analysis?
- Taking Advantage of Sparsity in Multi-Task Learning
- Guaranteed Minimum Rank Approximation from Linear Observations by Nuclear Norm Minimization with an Ellipsoidal Constraint
- Robust Matrix Decomposition with Outliers