Evading the curse of dimensionality in nonparametric density estimation with simplified vine copulas

Abstract

Practical applications of nonparametric density estimators in more than three dimensions suffer a great deal from the well-known curse of dimensionality: convergence slows down as dimension increases. We show that one can evade the curse of dimensionality by assuming a simplified vine copula model for the dependence between variables. We formulate a general nonparametric estimator for such a model and show under high-level assumptions that the speed of convergence is independent of dimension. We further discuss a particular implementation for which we validate the high-level assumptions and establish its asymptotic normality. Simulation experiments illustrate a large gain in finite sample performance when the simplifying assumption is at least approximately true. But even when it is severely violated, the vine copula based approach proves advantageous as soon as more than a few variables are involved. Lastly, we give an application of the estimator to a classification problem from astrophysics.

References in corpus (3)

• Parameter estimation for pair-copula constructions
• Semi-Supervised Domain Adaptation with Non-Parametric Copulas
• The partial vine copula: A dependence measure and approximation based on the simplifying assumption

Cited by in corpus (18)

• Prediction based on conditional distributions of vine copulas
• Nonparametric estimation of simplified vine copula models: comparison of methods
• Stochastic cosmic ray sources and the TeV break in the all-electron spectrum
• Information estimation using nonparametric copulas
• Estimating Non-Simplified Vine Copulas Using Penalized Splines
• Testing the simplifying assumption in high-dimensional vine copulas
• Copulas as High-Dimensional Generative Models: Vine Copula Autoencoders
• PRISMA: A Novel Approach for Deriving Probabilistic Surrogate Safety Measures for Risk Evaluation
• Measuring the Discrepancy between Conditional Distributions: Methods, Properties and Applications
• The partial vine copula: A dependence measure and approximation based on the simplifying assumption
• Understanding Convolutional Neural Networks with Information Theory: An Initial Exploration
• Doppler: Automated SKU Recommendation in Migrating SQL Workloads to the Cloud
• RoCourseNet: Distributionally Robust Training of a Prediction Aware Recourse Model
• Modeling high-dimensional dependence among astronomical data
• Dependence structure estimation using Copula Recursive Trees
• Alternative Approaches for Estimating Highest-Density Regions
• Semiparametric Copula Quantile Regression for Complete or Censored Data
• The bivariate $K$-finite normal mixture "blanket" copula