Regularization and variable selection via the elastic net

Summary.  We propose the elastic net, a new regularization and variable selection method. Real world data and a simulation study show that the elastic net often outperforms the lasso, while enjoying a similar sparsity of representation. In addition, the elastic net encourages a grouping effect, where strongly correlated predictors tend to be in or out of the model together. The elastic net is particularly useful when the number of predictors ( p ) is much bigger than the number of observations ( n ). By contrast, the lasso is not a very satisfactory variable selection method in the p ≫ n case. An algorithm called LARS-EN is proposed for computing elastic net regularization paths efficiently, much like algorithm LARS does for the lasso.     

On the grouped selection and model complexity of the adaptive elastic net

Lasso proved to be an extremely successful technique for simultaneous estimation and variable selection. However lasso has two major drawbacks. First, it does not enforce any grouping effect and secondly in some situation lasso solutions are inconsistent for variable selection. To overcome this inconsistency adaptive lasso is proposed where adaptive weights are used for penalizing different coefficients. Recently a doubly regularized technique namely elastic net is proposed which encourages grouping effect i.e. either selection or omission of the correlated variables together. However elastic net is also inconsistent. In this paper we study adaptive elastic net which does not have this drawback. In this article we specially focus on the grouped selection property of adaptive elastic net along with its model selection complexity. We also shed some light on the bias-variance tradeoff of different regularization methods including adaptive elastic net. An efficient algorithm was proposed in the line of LARS-EN, which is then illustrated with simulated as well as real life data examples.     

Adaptive Elastic Net GMM Estimation With Many Invalid Moment Conditions: Simultaneous Model and Moment Selection

This article develops the adaptive elastic net generalized method of moments (GMM) estimator in large-dimensional models with potentially (locally) invalid moment conditions, where both the number of structural parameters and the number of moment conditions may increase with the sample size. The basic idea is to conduct the standard GMM estimation combined with two penalty terms: the adaptively weighted lasso shrinkage and the quadratic regularization. It is a one-step procedure of valid moment condition selection, nonzero structural parameter selection (i.e., model selection), and consistent estimation of the nonzero parameters. The procedure achieves the standard GMM efficiency bound as if we know the valid moment conditions ex ante, for which the quadratic regularization is important. We also study the tuning parameter choice, with which we show that selection consistency still holds without assuming Gaussianity. We apply the new estimation procedure to dynamic panel data models, where both the time and cross-section dimensions are large. The new estimator is robust to possible serial correlations in the regression error terms.     

Gaussian Regularized Sliced Inverse Regression

Sliced Inverse Regression (SIR) is an effective method for dimension reduction in high-dimensional regression problems. The original method, however, requires the inversion of the predictors covariance matrix. In case of collinearity between these predictors or small sample sizes compared to the dimension, the inversion is not possible and a regularization technique has to be used. Our approach is based on a Fisher Lecture given by R.D. Cook where it is shown that SIR axes can be interpreted as solutions of an inverse regression problem. We propose to introduce a Gaussian prior distribution on the unknown parameters of the inverse regression problem in order to regularize their estimation. We show that some existing SIR regularizations can enter our framework, which permits a global understanding of these methods. Three new priors are proposed leading to new regularizations of the SIR method. A comparison on simulated data as well as an application to the estimation of Mars surface physical properties from hyperspectral images are provided.     

Bridge regression: Adaptivity and group selection

In high-dimensional regression problems regularization methods have been a popular choice to address variable selection and multicollinearity. In this paper we study bridge regression that adaptively selects the penalty order from data and produces flexible solutions in various settings. We implement bridge regression based on the local linear and quadratic approximations to circumvent the nonconvex optimization problem. Our numerical study shows that the proposed bridge estimators are a robust choice in various circumstances compared to other penalized regression methods such as the ridge, lasso, and elastic net. In addition, we propose group bridge estimators that select grouped variables and study their asymptotic properties when the number of covariates increases along with the sample size. These estimators are also applied to varying-coefficient models. Numerical examples show superior performances of the proposed group bridge estimators in comparisons with other existing methods.     

Selecting the regularization parameters in high-dimensional panel data models: Consistency and efficiency

This article considers panel data models in the presence of a large number of potential predictors and unobservable common factors. The model is estimated by the regularization method together with the principal components procedure. We propose a panel information criterion for selecting the regularization parameter and the number of common factors under a diverging number of predictors. Under the correct model specification, we show that the proposed criterion consistently identifies the true model. If the model is instead misspecified, the proposed criterion achieves asymptotically efficient model selection. Simulation results confirm these theoretical arguments.     

Adaptive elastic net-penalized quantile regression for variable selection

Abstract

There has been much attention on the high-dimensional linear regression models, which means the number of observations is much less than that of covariates. Considering the fact that the high dimensionality often induces the collinearity problem, in this article, we study the penalized quantile regression with the elastic net (EnetQR) that combines the strengths of the quadratic regularization and the lasso shrinkage. We investigate the weak oracle property of the EnetQR under mild conditions in the high dimensional setting. Moreover, we propose a two-step procedure, called adaptive elastic net quantile regression (AEnetQR), in which the weight vector in the second step is constructed from the EnetQR estimate in the first step. This two-step procedure is justified theoretically to possess the weak oracle property. The finite sample properties are performed through the Monte Carlo simulation and a real-data analysis.     

Grouping Variable Selection by Weight Fused Elastic Net for Multi-Collinear Data

In this article, we consider the problem of variable selection and estimation with the strongly correlated multi-collinear data by using grouping variable selection techniques. A new grouping variable selection method, called weight-fused elastic net(WFEN), is proposed to deal with the high dimensional collinear data. The proposed model, combined two different grouping effect mechanisms induced by the elastic net and weight-fused LASSO, respectively, can be easily unified in the frame of LASSO and computed efficiently. The performance with the simulation and real data sets shows that our method is competitive with other related methods, especially when the data present high multi-collinearity.     

Variable selection for generalized linear mixed models by L 1-penalized estimation

Generalized linear mixed models are a widely used tool for modeling longitudinal data. However, their use is typically restricted to few covariates, because the presence of many predictors yields unstable estimates. The presented approach to the fitting of generalized linear mixed models includes an L ₁-penalty term that enforces variable selection and shrinkage simultaneously. A gradient ascent algorithm is proposed that allows to maximize the penalized log-likelihood yielding models with reduced complexity. In contrast to common procedures it can be used in high-dimensional settings where a large number of potentially influential explanatory variables is available. The method is investigated in simulation studies and illustrated by use of real data sets.     

Functional variable selection via Gram–Schmidt orthogonalization for multiple functional linear regression

ABSTRACT

Functional linear model is of great practical importance, as exemplified by applications in high-throughput studies such as meteorological and biomedical research. In this paper, we propose a new functional variable selection procedure, called functional variable selection via Gram–Schmidt (FGS) orthogonalization, for a functional linear model with a scalar response and multiple functional predictors. Instead of the regularization methods, FGS takes into account the similarity between the functional predictors in a data-driven way and utilizes the technique of Gram–Schmidt orthogonalization to remove the irrelevant predictors. FGS can successfully discriminate between the relevant and the irrelevant functional predictors to achieve a high true positive ratio without including many irrelevant predictors, and yield explainable models, which offers a new perspective for the variable selection method in the functional linear model. Simulation studies are carried out to evaluate the finite sample performance of the proposed method, and also a weather data set is analysed.     

The Loss Rank Criterion for Variable Selection in Linear Regression Analysis

Abstract. Lasso and other regularization procedures are attractive methods for variable selection, subject to a proper choice of shrinkage parameter. Given a set of potential subsets produced by a regularization algorithm, a consistent model selection criterion is proposed to select the best one among this preselected set. The approach leads to a fast and efficient procedure for variable selection, especially in high‐dimensional settings. Model selection consistency of the suggested criterion is proven when the number of covariates d is fixed. Simulation studies suggest that the criterion still enjoys model selection consistency when d is much larger than the sample size. The simulations also show that our approach for variable selection works surprisingly well in comparison with existing competitors. The method is also applied to a real data set.     

Estimating a sparse reduction for general regression in high dimensions

Although the concept of sufficient dimension reduction that was originally proposed has been there for a long time, studies in the literature have largely focused on properties of estimators of dimension-reduction subspaces in the classical “small p, and large n” setting. Rather than the subspace, this paper considers directly the set of reduced predictors, which we believe are more relevant for subsequent analyses. A principled method is proposed for estimating a sparse reduction, which is based on a new, revised representation of an existing well-known method called the sliced inverse regression. A fast and efficient algorithm is developed for computing the estimator. The asymptotic behavior of the new method is studied when the number of predictors, p, exceeds the sample size, n, providing a guide for choosing the number of sufficient dimension-reduction predictors. Numerical results, including a simulation study and a cancer-drug-sensitivity data analysis, are presented to examine the performance.     

A descent algorithm for constrained LAD-Lasso estimation with applications in portfolio selection

To improve the out-of-sample performance of the portfolio, Lasso regularization is incorporated to the Mean Absolute Deviance (MAD)-based portfolio selection method. It is shown that such a portfolio selection problem can be reformulated as a constrained Least Absolute Deviance problem with linear equality constraints. Moreover, we propose a new descent algorithm based on the ideas of ‘nonsmooth optimality conditions’ and ‘basis descent direction set’. The resulting MAD-Lasso method enjoys at least two advantages. First, it does not involve the estimation of covariance matrix that is difficult particularly in the high-dimensional settings. Second, sparsity is encouraged. This means that assets with weights close to zero in the Markovwitz's portfolio are driven to zero automatically. This reduces the management cost of the portfolio. Extensive simulation and real data examples indicate that if the Lasso regularization is incorporated, MAD portfolio selection method is consistently improved in terms of out-of-sample performance, measured by Sharpe ratio and sparsity. Moreover, simulation results suggest that the proposed descent algorithm is more time-efficient than interior point method and ADMM algorithm.     

Robust sparse regression and tuning parameter selection via the efficient bootstrap information criteria

There is currently much discussion about lasso-type regularized regression which is a useful tool for simultaneous estimation and variable selection. Although the lasso-type regularization has several advantages in regression modelling, owing to its sparsity, it suffers from outliers because of using penalized least-squares methods. To overcome this issue, we propose a robust lasso-type estimation procedure that uses the robust criteria as the loss function, imposing L₁-type penalty called the elastic net. We also introduce to use the efficient bootstrap information criteria for choosing optimal regularization parameters and a constant in outlier detection. Simulation studies and real data analysis are given to examine the efficiency of the proposed robust sparse regression modelling. We observe that our modelling strategy performs well in the presence of outliers.     

Simultaneous estimation for semi-parametric multi-index models

Estimation of a general multi-index model comprises determining the number of linear combinations of predictors (structural dimension) that are related to the response, estimating the loadings of each index vector, selecting the active predictors and estimating the underlying link function. These objectives are often achieved sequentially at different stages of the estimation process. In this study, we propose a unified estimation approach under a semi-parametric model framework to attain these estimation goals simultaneously. The proposed estimation method is more efficient and stable than many existing methods where the estimation error in the structural dimension may propagate to the estimation of the index vectors and variable selection stages. A detailed algorithm is provided to implement the proposed method. Comprehensive simulations and a real data analysis illustrate the effectiveness of the proposed method.     

Random sliced inverse regression

Sliced Inverse Regression (SIR; 1991) is a dimension reduction method for reducing the dimension of the predictors without losing regression information. The implementation of SIR requires inverting the covariance matrix of the predictors—which has hindered its use to analyze high-dimensional data where the number of predictors exceed the sample size. We propose random sliced inverse regression (rSIR) by applying SIR to many bootstrap samples, each using a subset of randomly selected candidate predictors. The final rSIR estimate is obtained by aggregating these estimates. A simple variable selection procedure is also proposed using these bootstrap estimates. The performance of the proposed estimates is studied via extensive simulation. Application to a dataset concerning myocardial perfusion diagnosis from cardiac Single Proton Emission Computed Tomography (SPECT) images is presented.     

Independent screening in high-dimensional exponential family predictors’ space

We present a methodology for screening predictors that, given the response, follow a one-parameter exponential family distributions. Screening predictors can be an important step in regressions when the number of predictors p is excessively large or larger than n the number of observations. We consider instances where a large number of predictors are suspected irrelevant for having no information about the response. The proposed methodology helps remove these irrelevant predictors while capturing those linearly or nonlinearly related to the response.     

The Bayesian elastic net regression

A Bayesian elastic net approach is presented for variable selection and coefficient estimation in linear regression models. A simple Gibbs sampling algorithm was developed for posterior inference using a location-scale mixture representation of the Bayesian elastic net prior for the regression coefficients. The penalty parameters are chosen through an empirical method that maximizes the data marginal likelihood. Both simulated and real data examples show that the proposed method performs well in comparison to the other approaches.     

Lasso penalized semiparametric regression on high-dimensional recurrent event data via coordinate descent

This paper studies a fast computational algorithm for variable selection on high-dimensional recurrent event data. Based on the lasso penalized partial likelihood function for the response process of recurrent event data, a coordinate descent algorithm is used to accelerate the estimation of regression coefficients. This algorithm is capable of selecting important predictors for underdetermined problems where the number of predictors far exceeds the number of cases. The selection strength is controlled by a tuning constant that is determined by a generalized cross-validation method. Our numerical experiments on simulated and real data demonstrate the good performance of penalized regression in model building for recurrent event data in high-dimensional settings.     

Model-Free Feature Screening for Ultrahigh Dimensional Data

With the recent explosion of scientific data of unprecedented size and complexity, feature ranking and screening are playing an increasingly important role in many scientific studies. In this article, we propose a novel feature screening procedure under a unified model framework, which covers a wide variety of commonly used parametric and semiparametric models. The new method does not require imposing a specific model structure on regression functions, and thus is particularly appealing to ultrahigh-dimensional regressions, where there are a huge number of candidate predictors but little information about the actual model forms. We demonstrate that, with the number of predictors growing at an exponential rate of the sample size, the proposed procedure possesses consistency in ranking, which is both useful in its own right and can lead to consistency in selection. The new procedure is computationally efficient and simple, and exhibits a competent empirical performance in our intensive simulations and real data analysis.