Asymptotic Errors for High-Dimensional Convex Penalized Linear Regression beyond Gaussian Matrices.
Cédric GerbelotAlia AbbaraFlorent KrzakalaPublished in: COLT (2020)
Keyphrases
- linear regression
- least squares
- high dimensional
- doubly stochastic
- generalized linear models
- singular value decomposition
- maximum likelihood
- variable selection
- covariance matrices
- regression methods
- linear regression model
- regression problems
- low dimensional
- loss bounds
- optical flow
- dimension reduction
- ridge regression
- linear models
- gaussian mixture model
- feature space
- convex optimization
- locally weighted
- dimensionality reduction
- nonlinear regression
- multivariate regression
- convex hull
- decision trees
- regression trees
- linear model
- high dimensional data
- kernel function
- worst case
- pairwise