A two-stage Bayesian semiparametric model for novelty detection with robust prior information

DOI10.1007/S11222-021-10017-7zbMATH Open1475.62027arXiv2006.09012OpenAlexW3034724237MaRDI QIDQ2058768FDOQ2058768

Andrea Cappozzo, Francesca Greselin, Francesco D'Angelo

Publication date: 9 December 2021

Published in: Statistics and Computing (Search for Journal in Brave)

Abstract: Novelty detection methods aim at partitioning the test units into already observed and previously unseen patterns. However, two significant issues arise: there may be considerable interest in identifying specific structures within the novelty, and contamination in the known classes could completely blur the actual separation between manifest and new groups. Motivated by these problems, we propose a two-stage Bayesian semiparametric novelty detector, building upon prior information robustly extracted from a set of complete learning units. We devise a general-purpose multivariate methodology that we also extend to handle functional data objects. We provide insights on the model behavior by investigating the theoretical properties of the associated semiparametric prior. From the computational point of view, we propose a suitable $$-sequence to construct an independent slice-efficient sampler that takes into account the difference between manifest and novelty components. We showcase our model performance through an extensive simulation study and applications on both multivariate and functional datasets, in which diverse and distinctive unknown patterns are discovered.

Full work available at URL: https://arxiv.org/abs/2006.09012

zbMATH Keywords

Bayesian mixture modelBayesian nonparametricsnovelty detectionslice samplerminimum regularized covariance determinant

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08) Bayesian inference (62F15) Nonparametric estimation (62G05)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• The minimum regularized covariance determinant estimator
• Asymptotic Behaviour of the Posterior Distribution in Overfitted Mixture Models
• Functional data analysis.
• Bayesian Density Estimation and Inference Using Mixtures
• A practical guide to splines.
• Model-based clustering based on sparse finite Gaussian mixtures
• Asymptotics for the minimum covariance determinant estimator
• Bayesian cluster analysis: point estimation and credible balls (with discussion)
• Influence function and efficiency of the minimum covariance determinant scatter matrix estimator
• A Bayesian analysis of some nonparametric problems
• Least Median of Squares Regression
• Bayesian cluster analysis
• Central limit theorem and influence function for the MCD estimators at general multivariate distributions
• Fast and robust discriminant analysis
• Comparing clusterings -- an information based distance
• The two-parameter Poisson-Dirichlet distribution derived from a stable subordinator
• Exchangeable and partially exchangeable random partitions
• Slice sampling mixture models
• Gibbs Sampling Methods for Stick-Breaking Priors
• On a class of Bayesian nonparametric estimates: I. Density estimates
• Best approximations to random variables based on trimming procedures
• Hybrid Dirichlet mixture models for functional data
• Unsupervised Curve Clustering using B‐Splines
• Bayesian Semiparametric Joint Models for Functional Predictors
• Adaptive mixture discriminant analysis for supervised learning with unobserved classes
• Bayesian nonparametric classification for spectroscopy data
• Functional clustering in nested designs: modeling variability in reproductive epidemiology studies
• Variable selection and updating in model-based discriminant analysis for high dimensional data with food authenticity applications
• Robust and efficient estimation of multivariate scatter and location
• Semiparametric Bayes hierarchical models with mean and variance constraints
• On the Pitman–Yor process with spike and slab base measure
• Bayesian Hierarchical Functional Data Analysis Via Contaminated Informative Priors
• Anomaly and novelty detection for robust semi-supervised learning
• On the inferential implications of decreasing weight structures in mixture models

Cited In (1)

• Variational inference for semiparametric Bayesian novelty detection in large datasets

This page was built for publication: A two-stage Bayesian semiparametric model for novelty detection with robust prior information

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2058768)