Alexandre Belloni

Journal Article Operations Research · March 1, 2024 We consider a platform that serves (observable) agents, who belong to a larger network that also includes additional agents who are not served by the platform. We refer to the latter group of agents as latent agents. Associated with each agent are the agen ... Full text Cite

HIGH-DIMENSIONAL LATENT PANEL QUANTILE REGRESSION WITH AN APPLICATION TO ASSET PRICING

Journal Article Annals of Statistics · February 1, 2023 We propose a generalization of the linear panel quantile regression model to accommodate both sparse and dense parts: sparse means that while the number of covariates available is large, potentially only a much smaller number of them have a nonzero impact ... Full text Cite

High-dimensional linear models with many endogenous variables

Journal Article Journal of Econometrics · May 1, 2022 High-dimensional linear models with endogenous variables play an increasingly important role in the recent econometric literature. In this work, we allow for models with many endogenous variables and make use of many instrumental variables to achieve ident ... Full text Cite

Conditional quantile processes based on series or many regressors

Journal Article Journal of Econometrics · November 1, 2019 Quantile regression (QR) is a principal regression method for analyzing the impact of covariates on outcomes. The impact is described by the conditional quantile function and its functionals. In this paper we develop the nonparametric QR-series framework, ... Full text Cite

Valid Post-Selection Inference in High-Dimensional Approximately Sparse Quantile Regression Models

Journal Article Journal of the American Statistical Association · April 3, 2019 This work proposes new inference methods for a regression coefficient of interest in a (heterogenous) quantile regression model. We consider a high-dimensional model where the number of regressors potentially exceeds the sample size but a subset of them su ... Full text Cite

UNIFORMLY VALID POST-REGULARIZATION CONFIDENCE REGIONS FOR MANY FUNCTIONAL PARAMETERS IN Z-ESTIMATION FRAMEWORK.

Journal Article Annals of statistics · December 2018 In this paper, we develop procedures to construct simultaneous confidence bands for p ˜ potentially infinite-dimensional parameters after model selection for general moment condition models where p ˜ is potentially much larger than the sample size of a ... Full text Cite

Resource allocation under demand uncertainty and private information

Journal Article Management Science · December 1, 2017 We study the effect of multilateral private information on the efficiency of markets where capacity-constrained upstream agents supply a resource to downstream entities facing uncertain end-demands. We analyze two models: a "pooling system," in which a sin ... Full text Cite

Linear and conic programming estimators in high dimensional errors-in-variables models

Journal Article Journal of the Royal Statistical Society. Series B: Statistical Methodology · June 1, 2017 We consider the linear regression model with observation error in the design. In this setting, we allow the number of covariates to be much larger than the sample size. Several new estimation methods have been recently introduced for this model. Indeed, th ... Full text Cite

Mechanism and network design with private negative externalities

Journal Article Operations Research · May 1, 2017 A revenue-maximizing monopolist is selling a single indivisible good to buyers who face a loss if any of its rival buyers obtain it. The rivalry is modeled through a network, an arc between a pair of buyers indicates that a buyer considers another buyer it ... Full text Cite

Approximate group context tree

Journal Article Annals of Statistics · February 1, 2017 We study a variable length Markov chain model associated with a group of stationary processes that share the same context tree but each process has potentially different conditional probabilities. We propose a new model selection and estimation method whic ... Full text Cite

Program Evaluation and Causal Inference With High-Dimensional Data

Journal Article Econometrica · January 1, 2017 In this paper, we provide efficient estimators and honest confidence bands for a variety of treatment effects including local average (LATE) and local quantile treatment effects (LQTE) in data-rich environments. We can handle very many control variables, e ... Full text Cite

High-dimensional quantile regression

Chapter · January 1, 2017 High-dimensional models arise from the need for practitioners to improve the accuracy and validity of current models and to handle the increasing availability of data. Large models can arise from using a very flexible specification with many parameters whe ... Full text Cite

Post-Selection Inference for Generalized Linear Models With Many Controls

Journal Article Journal of Business and Economic Statistics · October 1, 2016 This article considers generalized linear models in the presence of many controls. We lay out a general methodology to estimate an effect of interest based on the construction of an instrument that immunizes against model selection mistakes and apply it to ... Full text Cite

Inference in High-Dimensional Panel Models With an Application to Gun Control

Journal Article Journal of Business and Economic Statistics · October 1, 2016 We consider estimation and inference in panel data models with additive unobserved individual specific heterogeneity in a high-dimensional setting. The setting allows the number of time-varying regressors to be larger than the sample size. To make informat ... Full text Cite

An {ℓ1, ℓ2, ℓ∞}-regularization approach to high-dimensional errors-in-variables models

Journal Article Electronic Journal of Statistics · January 1, 2016 Several new estimation methods have been recently proposed for the linear regression model with observation errors in the design. Different assumptions on the data generating process have motivated different estimators and analysis. In particular, the lite ... Full text Cite

Quantreg.nonpar: An r package for performing nonparametric series quantile regression

Journal Article R Journal · January 1, 2016 The R package quantreg.nonpar implements nonparametric quantile regression methods to estimate and make inference on partially linear quantile models. quantreg.nonpar obtains point estimates of the conditional quantile function and its derivatives based on ... Full text Cite

Comment

Journal Article Journal of the American Statistical Association · October 2, 2015 Full text Cite

Some new asymptotic theory for least squares series: Pointwise and uniform results

Journal Article Journal of Econometrics · June 1, 2015 In econometric applications it is common that the exact form of a conditional expectation is unknown and having flexible functional forms can lead to improvements over a pre-specified functional form, especially if they nest some successful parametric econ ... Full text Cite

Uniform post-selection inference for least absolute deviation regression and other Z-estimation problems

Journal Article Biometrika · March 1, 2015 We develop uniformly valid confidence regions for regression coefficients in a highdimensional sparse median regression model with homoscedastic errors. Our methods are based on amoment equation that is immunized against nonregular estimation of the nuisan ... Full text Cite

Escaping the local minima via simulated annealing: Optimization of approximately convex functions

Conference Journal of Machine Learning Research · January 1, 2015 We consider the problem of optimizing an approximately convex function over a bounded convex set in Rn using only function evaluations. The problem is reduced to sampling from an approximately log-concave distribution using the Hit-and-Run method, which is ... Cite

Latent Agents in Networks: Estimation and Targeting

HIGH-DIMENSIONAL LATENT PANEL QUANTILE REGRESSION WITH AN APPLICATION TO ASSET PRICING

High-dimensional linear models with many endogenous variables

Conditional quantile processes based on series or many regressors

Valid Post-Selection Inference in High-Dimensional Approximately Sparse Quantile Regression Models

UNIFORMLY VALID POST-REGULARIZATION CONFIDENCE REGIONS FOR MANY FUNCTIONAL PARAMETERS IN Z-ESTIMATION FRAMEWORK.

Resource allocation under demand uncertainty and private information

Linear and conic programming estimators in high dimensional errors-in-variables models

Mechanism and network design with private negative externalities

Approximate group context tree

Program Evaluation and Causal Inference With High-Dimensional Data

High-dimensional quantile regression

Post-Selection Inference for Generalized Linear Models With Many Controls

Inference in High-Dimensional Panel Models With an Application to Gun Control

An {ℓ1, ℓ2, ℓ∞}-regularization approach to high-dimensional errors-in-variables models

Quantreg.nonpar: An r package for performing nonparametric series quantile regression

Comment

Journal Article Journal of the American Statistical Association · October 2, 2015 Full text Cite

Some new asymptotic theory for least squares series: Pointwise and uniform results

Uniform post-selection inference for least absolute deviation regression and other Z-estimation problems

Escaping the local minima via simulated annealing: Optimization of approximately convex functions

High-dimensional methods and inference on structural and treatment effects

Journal Article Journal of Economic Perspectives · January 1, 2014 Full text Cite

Posterior inference in curved exponential families under increasing dimensions

Journal Article Econometrics Journal · January 1, 2014 Summary: In this paper, we study the large-sample properties of the posterior-based inference in the curved exponential family under increasing dimensions. The curved structure arises from the imposition of various restrictions on the model, such as moment ... Full text Cite

Pivotal estimation via square-root lasso in nonparametric regression

Journal Article Annals of Statistics · January 1, 2014 We propose a self-tuning √ Lasso method that simultaneously resolves three important practical problems in high-dimensional regression analysis, namely it handles the unknown scale, heteroscedasticity and (drastic) non- Gaussianity of the noise. In additio ... Full text Cite

Least squares after model selection in high-dimensional sparse models

Journal Article Bernoulli · January 1, 2013 In this article we study post-model selection estimators that apply ordinary least squares (OLS) to the model selected by first-step penalized estimators, typically Lasso. It is well known that Lasso can estimate the nonparametric regression function at ne ... Full text Cite

Inference on treatment effects after selection among high-dimensional controls

Journal Article Review of Economic Studies · January 1, 2013 We propose robust methods for inference about the effect of a treatment variable on a scalar outcome in the presence of very many regressors in a model with possibly non-Gaussian and heteroscedastic disturbances. We allow for the number of regressors to be ... Full text Cite

Sparse Models and Methods for Optimal Instruments With an Application to Eminent Domain

Journal Article Econometrica · November 1, 2012 We develop results for the use of Lasso and post-Lasso methods to form first-stage predictions and estimate optimal instruments in linear instrumental variables (IV) models with many instruments, p. Our results apply even when p is much larger than the sam ... Full text Cite

Optimal admission and scholarship decisions: Choosing customized marketing offers to attract a desirable mix of customers

Journal Article Marketing Science · July 1, 2012 Each year in the postsecondary education industry, schools offer admission to nearly 3 million new students and scholarships totaling nearly $100 billion. This is a large, understudied targeted marketing and price discrimination problem. This problem falls ... Full text Cite

Square-root lasso: Pivotal recovery of sparse signals via conic programming

Journal Article Biometrika · December 1, 2011 We propose a pivotal method for estimating high-dimensional sparse linear regression models, where the overall number of regressors p is large, possibly much larger than n, but only s regressors are significant. The method is a modification of the lasso, c ... Full text Cite

High Dimensional Sparse Econometric Models: An Introduction

Journal Article MIT Department of Economics Working Paper · June 26, 2011 Cite

On multivariate quantiles under partial orders

Journal Article The Annals of Statistics · April 1, 2011 Full text Cite

Ell;1-penalized ruantile regression in high-himensional sparse models

Journal Article Annals of Statistics · February 1, 2011 We consider median regression and, more generally, a possibly infinite collection of quantile regressions in high-dimensional sparse models. In these models, the number of regressors p is very large, possibly larger than the sample size n, but only at most ... Full text Cite

Inference for high-dimensional sparse econometric models

Journal Article · January 1, 2011 Introduction We consider linear, high-dimensional sparse (HDS) regression models in econometrics. The HDS regression model allows for a large number of regressors, p, which is possibly much larger than the sample size, n, but imposes that the model is spar ... Full text Cite

Multidimensional mechanism design: Finite-dimensional approximations and efficient computation

Journal Article Operations Research · July 1, 2010 Multidimensional mechanism design problems have proven difficult to solve by extending techniques from the onedimensional case. This paper considers mechanism design problems with multidimensional types when the seller's cost function is not separable acro ... Full text Open Access Cite

Dynamic bundle methods

Journal Article Mathematical Programming · September 1, 2009 Lagrangian relaxation is a popular technique to solve difficult optimization problems. However, the applicability of this technique depends on having a relatively low number of hard constraints to dualize. When there are many hard constraints, it may be pr ... Full text Cite

An Efficient Rescaled Perceptron Algorithm for Conic Systems

Journal Article Mathematics of Operations Research · August 1, 2009 The classical perceptron algorithm is an elementary row-action/relaxation algorithm for solving a homogeneous linear inequality system Ax > 0. A natural condition measure associated with this algorithm is the Euclidean width t of the cone of feasible solut ... Full text Cite

On the computational complexity of MCMC-based estimators in large samples

Journal Article Annals of Statistics · August 1, 2009 In this paper we examine the implications of the statistical large sample theory for the computational complexity of Bayesian and quasi-Bayesian estimation carried out using Metropolis random walks. Our analysis is motivated by the Laplace-Bernstein-Von Mi ... Full text Cite

A geometric analysis of Renegar's condition number, and its interplay with conic curvature

Journal Article Mathematical Programming · June 1, 2009 For a conic linear system of the form Ax K, K a convex cone, several condition measures have been extensively studied in the last dozen years. Among these, Renegar's condition number {\mathcal{C}}(A) is arguably the most prominent for its relation to data ... Full text Cite

Projective re-normalization for improving the behavior of a homogeneous conic linear system

Journal Article Mathematical Programming · May 1, 2009 In this paper we study the homogeneous conic system F: Ax = 0, x in C setminus 0 . We choose a point in ∫ C that serves as a normalizer and consider computational properties of the normalized system Fs : Ax = 0, bar s-T x = 1, x ∈ C . We show that the comp ... Full text Cite

On the second-order feasibility cone: Primal-dual representation and efficient projection

Journal Article SIAM Journal on Optimization · December 1, 2008 We study the second-order feasibility cone F = {y ∈ ℝn : ∥ My ∥ ≤ gTy} for given data (M,g). We construct a new representation for this cone and its dual based on the spectral decomposition of the matrix MTM - ggT. This representation is used to efficientl ... Full text Cite

On the Behrens-Fisher Problem: A globally convergent algorithm and a finite-sample study of the wald, LR and LM tests

Journal Article Annals of Statistics · October 1, 2008 In this paper we provide a provably convergent algorithm for the multivariate Gaussian Maximum Likelihood version of the Behrens-Fisher Problem. Our work builds upon a formulation of the log-likelihood function proposed by Buot and Richards [5]. Instead of ... Full text Cite

Optimizing product line designs: Efficient methods and comparisons

Journal Article Management Science · September 1, 2008 We take advantage of recent advances in optimization methods and computer hardware to identify globally optimal solutions of product line design problems that are too large for complete enumeration. We then use this guarantee of global optimality to benchm ... Full text Cite

Norm-induced densities and testing the boundedness of a convex set

Journal Article Mathematics of Operations Research · February 1, 2008 In this paper, we explore properties of a family of probability density functions, called norm-induced densities, defined as f t(x)={e -t||x||p/∫ Ke -t||y||pdy, x ∈ K 0, x ∉ K, where K is a n-dimensional convex set that contains the origin, parameters t > ... Full text Cite

On the symmetry function of a convex set

Journal Article Mathematical Programming · January 1, 2008 We attempt a broad exploration of properties and connections between the symmetry function of a convex set S ⊂ ℝn and other arenas of convexity including convex functions, convex geometry, probability theory on convex sets, and computational complexity. Gi ... Full text Cite

An efficient re-scaled perceptron algorithm for conic systems

Journal Article Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) · January 1, 2007 The classical perceptron algorithm is an elementary algorithm for solving a homogeneous linear inequality system Ax > 0, with many important applications in learning theory (e.g., [11,8]). A natural condition measure associated with this algorithm is the E ... Full text Cite

Lagrangian Heuristics for the Linear Ordering Problem

Chapter · 2004 Cite

Bundle Relaxation and Primal Recovery in Unit Commitment Problems. The Brazilian Case

Journal Article Annals of Operations Research · April 1, 2003 We consider the inclusion of commitment of thermal generation units in the optimal management of the Brazilian power system. By means of Lagrangian relaxation we decompose the problem and obtain a nondifferentiable dual function that is separable. We solve ... Full text Cite