Robust Wasserstein profile inference and applications to machine learning

DOI10.1017/jpr.2019.49zbMath1436.62336arXiv1610.05627OpenAlexW2537619949WikidataQ92196462 ScholiaQ92196462MaRDI QIDQ5235055

Karthyek R. A. Murthy, Yang Kang, Jose H. Blanchet

Publication date: 7 October 2019

Published in: Journal of Applied Probability (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1610.05627

zbMATH Keywords

regularizationlogistic regressionempirical likelihoodWasserstein distancesupport vector machinedistributionally robust optimizationsquare-root Lassolimit characterization of optimal Wasserstein ball radius and regularization parameter

Mathematics Subject Classification ID

Ridge regression; shrinkage estimators (Lasso) (62J07) Generalized linear models (logistic models) (62J12) Learning and adaptive systems in artificial intelligence (68T05)

Related Items (35)

Learning models with uniform performance via distributionally robust optimization ⋮ Regularization via Mass Transportation ⋮ Gromov–Wasserstein distances between Gaussian distributions ⋮ Frameworks and results in distributionally robust optimization ⋮ On the regularized risk of distributionally robust learning over deep neural networks ⋮ Tractable reformulations of two-stage distributionally robust linear programs over the type-\(\infty\) Wasserstein ball ⋮ Optimal Transport-Based Distributionally Robust Optimization: Structural Properties and Iterative Schemes ⋮ Distributionally robust profit opportunities ⋮ Optimizing decisions for a dual-channel retailer with service level requirements and demand uncertainties: a Wasserstein metric-based distributionally robust optimization approach ⋮ Distributionally Robust Linear and Discrete Optimization with Marginals ⋮ Distributions with maximum spread subject to Wasserstein distance constraints ⋮ On approximations of data-driven chance constrained programs over Wasserstein balls ⋮ Distributionally robust chance constrained games under Wasserstein ball ⋮ Bayesian Distributionally Robust Optimization ⋮ Unnamed Item ⋮ Distributionally robust portfolio optimization with second-order stochastic dominance based on Wasserstein metric ⋮ An Improved Central Limit Theorem and Fast Convergence Rates for Entropic Transportation Costs ⋮ Regularization for Wasserstein distributionally robust optimization ⋮ Data-driven distributionally robust risk-averse two-stage stochastic linear programming over Wasserstein ball ⋮ Markov decision processes under model uncertainty ⋮ Solving multistage stochastic linear programming via regularized linear decision rules: an application to hydrothermal dispatch planning ⋮ On distributionally robust chance constrained programs with Wasserstein distance ⋮ Scalable Algorithms for the Sparse Ridge Regression ⋮ Robust arbitrage conditions for financial markets ⋮ Wasserstein distributionally robust chance-constrained optimization for energy and reserve dispatch: an exact and physically-bounded formulation ⋮ Beyond Gaussian approximation: bootstrap for maxima of sums of independent random vectors ⋮ Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach ⋮ A data-driven approach for a class of stochastic dynamic optimization problems ⋮ Robust Wasserstein profile inference and applications to machine learning ⋮ Distributionally robust optimization. A review on theory and applications ⋮ Dynamics of Data-driven Ambiguity Sets for Hyperbolic Conservation Laws with Uncertain Inputs ⋮ Sample Out-of-Sample Inference Based on Wasserstein Distance ⋮ Distributionally robust bottleneck combinatorial problems: uncertainty quantification and robust decision making ⋮ Formulation and properties of a divergence used to compare probability measures without absolute continuity ⋮ Tight bounds for a class of data-driven distributionally robust risk measures

Uses Software

• flare

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• On the rate of convergence in Wasserstein distance of the empirical measure
• Characterization of the equivalence of robustification and regularization in linear and matrix regression
• Empirical likelihood ratio confidence regions
• Extending the scope of empirical likelihood
• Empirical likelihood for linear models
• Matching random samples in many dimensions
• Smoothed empirical likelihood confidence intervals for quantiles
• Empirical likelihood and general estimating equations
• Mass transportation problems. Vol. 1: Theory. Vol. 2: Applications
• Weighted empirical likelihood inference.
• The empirical likelihood approach to quantifying uncertainty in sample average approximation
• Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations
• Asymptotics for Lasso-type estimators.
• The earth mover's distance as a metric for image retrieval
• Least angle regression. (With discussion)
• Simultaneous analysis of Lasso and Dantzig selector
• The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
• On sharpness of Tchebycheff-type inequalities
• Square-root lasso: pivotal recovery of sparse signals via conic programming
• Empirical likelihood ratio confidence intervals for a single functional
• Earth mover's distances on discrete surfaces
• Generalized Chebychev Inequalities: Theory and Applications in Decision Analysis
• Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach
• Sample Out-of-Sample Inference Based on Wasserstein Distance
• Calibration of Distributionally Robust Empirical Optimization Models
• Recovering Best Statistical Guarantees via the Empirical Divergence-Based Distributionally Robust Optimization
• Regularization via Mass Transportation
• Quantifying Distributional Model Risk via Optimal Transport
• Robust Wasserstein profile inference and applications to machine learning
• Robust Regression and Lasso
• EMPIRICAL LIKELIHOOD BASED INFERENCE WITH APPLICATIONS TO SOME ECONOMETRIC MODELS
• Optimal Transport
• A unified framework for high-dimensional analysis of \(M\)-estimators with decomposable regularizers

This page was built for publication: Robust Wasserstein profile inference and applications to machine learning