Support recovery without incoherence: A case for nonconvex regularization
نویسندگان
چکیده
منابع مشابه
Support recovery without incoherence: A case for nonconvex regularization
We develop a new primal-dual witness proof framework that may be used to establish variable selection consistency and ∞-bounds for sparse regression problems, even when the loss function and regularizer are nonconvex. We use this method to prove two theorems concerning support recovery and ∞-guarantees for a regression estimator in a general setting. Notably, our theory applies to all potential...
متن کاملNonconvex Regularization for Image Segmentation
We propose a new method for image segmentation based on a variational regularization algorithm for image denoising. We modify the Rudin-Osher-Fatemi (ROF) model in [1] by minimizing the p L -norm of the gradient, where 0 > p is very small. The result is that we better preserve edges, while flattening regions away from the edges. This results in an automatic segmentation of the image into severa...
متن کاملStochastic Cubic Regularization for Fast Nonconvex Optimization
This paper proposes a stochastic variant of a classic algorithm—the cubic-regularized Newton method [Nesterov and Polyak, 2006]. The proposed algorithm efficiently escapes saddle points and finds approximate local minima for general smooth, nonconvex functions in only Õ( −3.5) stochastic gradient and stochastic Hessian-vector product evaluations. The latter can be computed as efficiently as sto...
متن کاملNeutrality without incoherence: a response to Clark.
In an earlier paper [1], we found that equations based on the neutral theory of biodiversity enjoyed considerable success in predicting the relaxation times for avifaunal extinctions on isolated habitats. We suggested that the neutral theory of biodiversity might turn out to be a useful tool in conservation biology. However, Clark [2] maintains that our study ‘does not need UNTB, neither does i...
متن کاملA Graduated Nonconvex Regularization for Sparse High Dimensional Model Estimation
Many high dimensional data mining problems can be formulated as minimizing an empirical loss function with a penalty proportional to the number of variables required to describe a model. We propose a graduated non-convexification method to facilitate tracking of a global minimizer of this problem. We prove that under some conditions the proposed regularization problem using the continuous piece...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: The Annals of Statistics
سال: 2017
ISSN: 0090-5364
DOI: 10.1214/16-aos1530