Fisher Lecture: Dimension Reduction in Regression

Abstract

Beginning with a discussion of R. A. Fisher's early written remarks that relate to dimension reduction, this article revisits principal components as a reductive method in regression, develops several model-based extensions and ends with descriptions of general approaches to model-based and model-free dimension reduction in regression. It is argued that the role for principal components and related methodology may be broader than previously seen and that the common practice of conditioning on observed values of the predictors may unnecessarily limit the choice of regression methodology.

This paper commented in: [arXiv:0708.3776], [arXiv:0708.3777], [arXiv:0708.3779]. Rejoinder in [arXiv:0708.3781]. Published at http://dx.doi.org/10.1214/088342306000000682 in the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

Cited by in corpus (55)

• Deep Learning for Short-Term Traffic Flow Prediction
• Coordinate-independent sparse sufficient dimension reduction and variable selection
• Deep Learning: A Bayesian Perspective
• Principal support vector machines for linear and nonlinear sufficient dimension reduction
• Principal Fitted Components for Dimension Reduction in Regression
• Dimension reduction for nonelliptically distributed predictors
• A general theory for nonlinear sufficient dimension reduction: Formulation and estimation
• High-Dimensional Regression with Gaussian Mixtures and Partially-Latent Response Variables
• Distributed multinomial regression
• HI Content and Optical Properties of Field Galaxies from the ALFALFA Survey. II. Multivariate Analysis of a Galaxy Sample in Low Density Environments
• Co-Localization of Audio Sources in Images Using Binaural Features and Locally-Linear Regression
• Deep Learning in Finance
• Estimating the Expected Value of Partial Perfect Information in Health Economic Evaluations using Integrated Nested Laplace Approximation
• A Selective Overview of Deep Learning
• Variable selection for general index models via sliced inverse regression
• Wavelet-domain regression and predictive inference in psychiatric neuroimaging
• Parametric or nonparametric? A parametricness index for model selection
• An adaptive composite quantile approach to dimension reduction
• Estimating sufficient reductions of the predictors in abundant high-dimensional regressions
• Penalized Orthogonal-Components Regression for Large p Small n Data
• On surrogate dimension reduction for measurement error regression: An invariance law
• Heteroscedastic BART Using Multiplicative Regression Trees
• Prediction regions through Inverse Regression
• Minimum Average Deviance Estimation for Sufficient Dimension Reduction
• Dimension reduction for Gaussian process emulation: an application to the influence of bathymetry on tsunami heights
• Deep Learning Partial Least Squares
• Deep Dimension Reduction for Supervised Representation Learning
• A Statistical Theory of Deep Learning via Proximal Splitting
• Theoretical properties of Cook's PFC dimension reduction algorithm for linear regression
• Predictor-dependent shrinkage for linear regression via partial factor modeling
• Randomized Dimension Reduction on Massive Data
• Multinomial Inverse Regression for Text Analysis
• Discussion of: Treelets--An adaptive multi-scale basis for sparse unordered data
• Sparse Generalized Eigenvalue Problem: Optimal Statistical Rates via Truncated Rayleigh Flow
• A Semiparametric Approach to Model Effect Modification
• Comment: Fisher Lecture: Dimension Reduction in Regression
• Nonlinear Level Set Learning for Function Approximation on Sparse Data with Applications to Parametric Differential Equations
• On the Impact of Dimension Reduction on Graphical Structures
• Adaptive Randomized Dimension Reduction on Massive Data
• Nonsparse learning with latent variables
• Sufficient Markov Decision Processes with Alternating Deep Neural Networks
• Sliced Inverse Regression for Spatial Data
• Certifiably Optimal Sparse Sufficient Dimension Reduction
• Merging Two Cultures: Deep and Statistical Learning
• On the explanatory power of principal components
• Flexible Principal Component Analysis for Exponential Family Distributions
• Generalized Matrix Decomposition Regression: Estimation and Inference for Two-way Structured Data
• Sufficient Dimension Reduction for High-Dimensional Regression and Low-Dimensional Embedding: Tutorial and Survey
• On expectile-assisted inverse regression estimation for sufficient dimension reduction
• Sufficient Dimension Reduction for Interactions
• Robust Variable and Interaction Selection for Logistic Regression and Multiple Index Models
• Supervised Multiscale Dimension Reduction for Spatial Interaction Networks
• Prediction analysis for microbiome sequencing data
• Gaussian Regularized Sliced Inverse Regression
• Sufficient Dimension Reduction for Populations with Structured Heterogeneity