Estimating conditional quantiles with the help of the pinball loss

Abstract

The so-called pinball loss for estimating conditional quantiles is a well-known tool in both statistics and machine learning. So far, however, only little work has been done to quantify the efficiency of this tool for nonparametric approaches. We fill this gap by establishing inequalities that describe how close approximate pinball risk minimizers are to the corresponding conditional quantile. These inequalities, which hold under mild assumptions on the data-generating distribution, are then used to establish so-called variance bounds, which recently turned out to play an important role in the statistical analysis of (regularized) empirical risk minimization approaches. Finally, we use both types of inequalities to establish an oracle inequality for support vector machines that use the pinball loss. The resulting learning rates are min--max optimal under some standard regularity assumptions on the conditional quantile.

Published in at http://dx.doi.org/10.3150/10-BEJ267 the Bernoulli (http://isi.cbs.nl/bernoulli/) by the International Statistical Institute/Bernoulli Society (http://isi.cbs.nl/BS/bshome.htm)

References in corpus (3)

• Local Rademacher complexities
• Statistical performance of support vector machines
• Regularization in kernel learning

Cited by in corpus (35)

• Conformalized Quantile Regression
• Calibrating AI Models for Few-Shot Demodulation via Conformal Prediction
• Accelerating Approximate Bayesian Computation with Quantile Regression: Application to Cosmological Redshift Distributions
• A general framework for multi-step ahead adaptive conformal heteroscedastic time series forecasting
• Robust Kernel-based Distribution Regression
• Improved conformalized quantile regression
• Probabilistic Multi-Step-Ahead Short-Term Water Demand Forecasting with Lasso
• Deep Quantile Regression: Mitigating the Curse of Dimensionality Through Composition
• High-dimensional sparse vine copula regression with application to genomic prediction
• A General Theory for Structured Prediction with Smooth Convex Surrogates
• A Bernstein-type Inequality for Some Mixing Processes and Dynamical Systems with an Application to Learning
• Improving Conditional Coverage via Orthogonal Quantile Regression
• With Malice Towards None: Assessing Uncertainty via Equalized Coverage
• Robust Nonparametric Regression with Deep Neural Networks
• Calibrated Multiple-Output Quantile Regression with Representation Learning
• Optimal Learning Rates for Localized SVMs
• Non-Asymptotic Analysis of Stochastic Approximation Algorithms for Streaming Data
• Probabilistic Forecasting in Day-Ahead Electricity Markets: Simulating Peak and Off-Peak Prices
• Ensemble Machine Learning Methods for Modeling COVID19 Deaths
• ML-AQP: Query-Driven Approximate Query Processing based on Machine Learning
• Mixed one-bit compressive sensing with applications to overexposure correction for CT reconstruction
• Asymptotic Confidence Sets for General Nonparametric Regression and Classification by Regularized Kernel Methods
• Generating Synthetic Rainfall Fields by R-vine Copulas Applied to Seamless Probabilistic Predictions
• Learning theory estimates with observations from general stationary stochastic processes
• Distribution-free uncertainty quantification for inverse problems: application to weak lensing mass mapping
• On the Energy Distribution of the Galactic Center Excess' Sources
• Universal Consistency and Robustness of Localized Support Vector Machines
• Direct Interval Propagation Methods using Neural-Network Surrogates for Uncertainty Quantification in Physical Systems Surrogate Model
• Achieving Skilled and Reliable Daily Probabilistic Forecasts of Wind Power at Subseasonal-to-Seasonal Timescales over France
• Navigating Time's Possibilities: Plausible Counterfactual Explanations for Multivariate Time-Series Forecast through Genetic Algorithms
• l1-norm quantile regression screening rule via the dual circumscribed sphere
• Learning Rates of Regression with q-norm Loss and Threshold
• Learning Rates for Kernel-Based Expectile Regression
• Learning rates for the risk of kernel based quantile regression estimators in additive models
• Estimation of scale functions to model heteroscedasticity by support vector machines