Flexible integrated functional depths

This paper develops a new class of functional depths. A generic member of this class is coined J th order k th moment integrated depth. It is based on the distribution of the cross-sectional halfspace depth of a function in the marginal evaluations (in time) of the random process. Asymptotic properties of the proposed depths are provided: we show that they are uniformly consistent and satisfy an inequality related to the law of the iterated logarithm. Moreover, limiting distributions are derived under mild regularity assumptions. The versatility displayed by the new class of depths makes them particularly amenable for capturing important features of functional distributions. This is illustrated in supervised learning, where we show that the corresponding maximum depth classifiers outperform classical competitors.

• Publication year

  2021

• Venue

  Bernoulli

• Publication date

  2021-02-01

• Fields of study

  Mathematics

• Identifiers
  DOI 10.3150/20-BEJ1254
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Computing the halfspace depth with multiple try algorithm and simulated annealing algorithm

  2019influential reference
• Data depth for measurable noisy random functions

  2019cited by this paper
• Weak convergence and empirical processes

  2019influential reference
• Depth-Based Recognition of Shape Outlying Functions

  2017cited by this paper
• Integrated depth for measurable functions and sets

  2017cited by this paper
• Functional Data Analysis

  2016cited by this paper
• Functional independent component analysis : an extension of fourth-order blind identification

  2016cited by this paper
• Integrated depth for functional data: statistical properties and consistency

  2016cited by this paper
• Depth and Depth-Based Classification with R Package ddalpha

  2016cited by this paper
• Consistency of non-integrated depths for functional data

  2015cited by this paper
• Extremal Depth for Functional Data and Applications

  2015influential reference
• The spatial distribution in infinite dimensional spaces and related quantiles and depths

  2014influential reference
• Fast Computation of Tukey Trimmed Regions and Median in Dimension p > 2

  2014cited by this paper
• Monge-Kantorovich Depth, Quantiles, Ranks and Signs

  2014cited by this paper
• Concerns with Functional Depth

  2013cited by this paper
• Spatial depth-based classification for functional data

  2013cited by this paper
• Sequential testing policies for complex systems under precedence constraints

  2013influential reference
• Depth Statistics

  2012cited by this paper
• DD-Classifier: Nonparametric Classification Procedure Based on DD-Plot

  2012influential reference
• General notions of depth for functional data

  2012cited by this paper
• Achieving near perfect classification for functional data

  2012influential reference
• A half-region depth for functional data

  2011influential reference
• Multivariate quantiles and multiple-output regression quantiles: from L1 optimization to halfspace depth

  2010cited by this paper
• On depth measures and dual statistics. A methodology for dealing with general data

  2009cited by this paper
• On the Concept of Depth for Functional Data

  2009influential reference
• Generation of synthetic sequences of half‐hourly temperature

  2008cited by this paper
• Robust estimation and classification for functional data via projection-based depth notions

  2007influential reference
• Nonparametric Functional Data Analysis: Theory And Practice

  2007cited by this paper
• Nearest neighbor classification in infinite dimension

  2006cited by this paper
• On Maximum Depth and Related Classifiers

  2005influential reference
• Functional classification in Hilbert spaces

  2005cited by this paper
• Asymptotics for the Tukey depth process, with an application to a multivariate trimmed mean

  2004influential reference
• Curves discrimination: a nonparametric functional approach

  2003cited by this paper
• Real Analysis and Probability: Measurability: Borel Isomorphism and Analytic Sets

  2002cited by this paper
• Trimmed means for functional data

  2001cited by this paper
• A Functional Data—Analytic Approach to Signal Discrimination

  2001cited by this paper
• Uniform Central Limit Theorems

  2001influential reference
• General notions of statistical depth function

  2000influential reference
• Multivariate analysis by data depth: descriptive statistics, graphics and inference, (with discussion and a rejoinder by Liu and Singh)

  1999cited by this paper
• Appendix: probability and measure

  1997cited by this paper
• Optimal minimal neural interpretation of spectra

  1992cited by this paper
• Breakdown Properties of Location Estimates Based on Halfspace Depth and Projected Outlyingness

  1992influential reference
• On a Notion of Data Depth Based on Random Simplices

  1990cited by this paper
• Rates of convergence in the central limit theorem for empirical processes

  1986cited by this paper
• Log Log Laws for Empirical Measures

  1980influential reference
• A limit theorem for measurable random processes and its applications

  1976cited by this paper
• Mathematics and the Picturing of Data

  1975cited by this paper

• Discriminating Tail Behavior Using Halfspace Depths: Population and Empirical Perspectives

  2025cites this paper
• Statistical Depth Meets Machine Learning: Kernel Mean Embeddings and Depth in Functional Data Analysis

  2021cites this paper
• Integrated shape-sensitive functional metrics

  2021cites this paper