QUADRO: a supervised dimension reduction method via Rayleigh quotient optimization
From MaRDI portal
Publication:2515488
Abstract: We propose a novel Rayleigh quotient based sparse quadratic dimension reduction method - named QUADRO (Quadratic Dimension Reduction via Rayleigh Optimization) - for analyzing high- dimensional data. Unlike in the linear setting where Rayleigh quotient optimization coincides with classification, these two problems are very different under nonlinear settings. In this paper, we clarify this difference and show that Rayleigh quotient optimization may be of independent scientific interests. One major challenge of Rayleigh quotient optimization is that the variance of quadratic statistics involves all fourth cross-moments of predictors, which are infeasible to compute for high-dimensional applications and may accumulate too many stochastic errors. This issue is resolved by considering a family of elliptical models. Moreover, for heavy-tail distributions, robust estimates of mean vectors and covariance matrices are employed to guarantee uniform convergence in estimating nonpolynomially many parameters, even though only the fourth moments are assumed. Methodologically, QUADRO is based on elliptical models which allow us to formulate the Rayleigh quotient maximization as a convex optimization problem. Computationally, we propose an efficient linearized augmented Lagrangian method to solve the constrained optimization problem. Theoretically, we provide explicit rates of convergence in terms of Rayleigh quotient under both Gaussian and general elliptical models. Thorough numerical results on both synthetic and real datasets are also provided to back up our theoretical results.
Recommendations
- Supervised dimensionality reduction via sequential semidefinite programming
- Sparse quadratic classification rules via linear dimension reduction
- Linear dimensionality reduction: survey, insights, and generalizations
- A convex optimization approach to high-dimensional sparse quadratic discriminant analysis
- Robust Rayleigh quotient minimization and nonlinear eigenvalue problems
Cites work
- A constrained \(\ell _{1}\) minimization approach to sparse precision matrix estimation
- A direct estimation approach to sparse linear discriminant analysis
- A NEW MEASURE OF RANK CORRELATION
- Challenging the empirical mean and empirical variance: a deviation study
- CODA: high dimensional copula discriminant analysis
- Comparison of sliced inverse regression approaches for underdetermined cases
- Coordinate-independent sparse sufficient dimension reduction and variable selection
- Estimation of Moment Parameter in Elliptical Distributions
- High-dimensional classification using features annealed independence rules
- High-dimensional semiparametric Gaussian copula graphical models
- scientific article; zbMATH DE number 5957408 (Why is no real title available?)
- scientific article; zbMATH DE number 3223982 (Why is no real title available?)
- On Directional Regression for Dimension Reduction
- One-step sparse estimates in nonconcave penalized likelihood models
- Penalized classification using Fisher's linear discriminant
- QUADRO: a supervised dimension reduction method via Rayleigh quotient optimization
- Regularized linear discriminant analysis and its application in microarrays
- Simultaneous analysis of Lasso and Dantzig selector
- Sliced Inverse Regression for Dimension Reduction
- Sliced Inverse Regression with Regularizations
- Sparse linear discriminant analysis by thresholding for high dimensional data
- Strong oracle optimality of folded concave penalized estimation
- The Adaptive Lasso and Its Oracle Properties
- Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
Cited in
(13)- Covariate-adjusted tensor classification in high dimensions
- Linear components of quadratic classifiers
- The robust nearest shrunken centroids classifier for high-dimensional heavy-tailed data
- Efficient computation for differential network analysis with applications to quadratic discriminant analysis
- Dynamic linear discriminant analysis in high dimensional space
- Large covariance estimation through elliptical factor models
- QUADRO: a supervised dimension reduction method via Rayleigh quotient optimization
- A direct approach for sparse quadratic discriminant analysis
- Linear discriminant analysis with sparse and dense signals
- Higher moment estimation for elliptically distributed data: is it necessary to use a sledgehammer to crack an egg?
- The Tucker low-rank classification model for tensor data
- High dimensional quadratic discriminant analysis: optimality and phase transitions
- Homogeneity Pursuit in Clustered Data Analysis When Cluster Sizes Are Small
This page was built for publication: QUADRO: a supervised dimension reduction method via Rayleigh quotient optimization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2515488)
