P
Peng Zhao
Researcher at University of California, Berkeley
Publications - 9
Citations - 4311
Peng Zhao is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Feature selection & Mean squared error. The author has an hindex of 6, co-authored 6 publications receiving 4088 citations.
Papers
More filters
Journal Article
On Model Selection Consistency of Lasso
Peng Zhao,Bin Yu +1 more
TL;DR: It is proved that a single condition, which is called the Irrepresentable Condition, is almost necessary and sufficient for Lasso to select the true model both in the classical fixed p setting and in the large p setting as the sample size n gets large.
Journal ArticleDOI
The composite absolute penalties family for grouped and hierarchical variable selection
Peng Zhao,Guilherme Rocha,Bin Yu +2 more
TL;DR: CAP is shown to improve on the predictive performance of the LASSO in a series of simulated experiments, including cases with $p\gg n$ and possibly mis-specified groupings, and iCAP is seen to be parsimonious in the experiments.
Journal ArticleDOI
The composite absolute penalties family for grouped and hierarchical variable selection
Peng Zhao,Guilherme Rocha,Bin Yu +2 more
TL;DR: In this paper, the Composite Absolute Penalties (CAP) family is proposed to combine different norms including L1 to form an intelligent penalty in order to add side information to the fitting of a regression or classification model to obtain reasonable estimates.
Grouped and Hierarchical Model Selection through Composite Absolute Penalties
Peng Zhao,Guilherme Rocha,Bin Yu +2 more
TL;DR: This paper introduces the Composite Absolute Penalties (CAP) family which allows the grouping and hierarchical relationships between the predictors to be expressed and combined norms including L1 to form an intelligent penalty in order to add side information to the fitting of a regression or classification model to obtain reasonable estimates.
ReportDOI
A path following algorithm for Sparse Pseudo-Likelihood Inverse Covariance Estimation (SPLICE)
Guilherme Rocha,Peng Zhao,Bin Yu +2 more
TL;DR: This paper proposes an ‘1 penalized pseudo-likelihood estimate for the inverse covariance matrix, and names it SPLICE, which gives the best overall performance in terms of three metrics on the precision matrix and ROC curve for model selection.