Neyman-Pearson classification: parametrics and sample size requirement

zbMATH Open1497.62159arXiv1802.02557MaRDI QIDQ4969044FDOQ4969044

Authors: Xin Tong, Lucy Xia, Jiacheng Wang, Yang Feng 

Publication date: 5 October 2020

Abstract: The Neyman-Pearson (NP) paradigm in binary classification seeks classifiers that achieve a minimal type II error while enforcing the prioritized type I error controlled under some user-specified level $alpha$. This paradigm serves naturally in applications such as severe disease diagnosis and spam detection, where people have clear priorities among the two error types. Recently, Tong, Feng and Li (2018) proposed a nonparametric umbrella algorithm that adapts all scoring-type classification methods (e.g., logistic regression, support vector machines, random forest) to respect the given type I error upper bound $alpha$ with high probability, without specific distributional assumptions on the features and the responses. Universal the umbrella algorithm is, it demands an explicit minimum sample size requirement on class $0$, which is often the more scarce class, such as in rare disease diagnosis applications. In this work, we employ the parametric linear discriminant analysis (LDA) model and propose a new parametric thresholding algorithm, which does not need the minimum sample size requirements on class $0$ observations and thus is suitable for small sample applications such as rare disease diagnosis. Leveraging both the existing nonparametric and the newly proposed parametric thresholding rules, we propose four LDA-based NP classifiers, for both low- and high-dimensional settings. On the theoretical front, we prove NP oracle inequalities for one proposed classifier, where the rate for excess type II error benefits from the explicit parametric model assumption. Furthermore, as NP classifiers involve a sample splitting step of class $0$ observations, we construct a new adaptive sample splitting scheme that can be applied universally to NP classifiers, and this adaptive strategy reduces the type II error of these classifiers.

Full work available at URL: https://arxiv.org/abs/1802.02557

Recommendations

• Neyman-Pearson classification under high-dimensional settings
• A plug-in approach to Neyman-Pearson classification
• Neyman-Pearson classification, convexity and stochastic constraints
• A Neyman–Pearson Approach to Statistical Learning
• Analysis to Neyman-Pearson classification with convex loss function

zbMATH Keywords

classificationadaptive splittinglinear discriminant analysis (LDA)asymmetric errorNeyman-Pearson (NP) paradigmminimum sample size requirementNP oracle inequalitiesNP umbrella algorithm

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Learning and adaptive systems in artificial intelligence (68T05)

Cites Work

• Regularized linear discriminant analysis and its application in microarrays
• Penalized classification using Fisher's linear discriminant
• Random forests
• Title not available (Why is that?)
• A direct approach to sparse discriminant analysis in ultra-high dimensions
• A direct estimation approach to sparse linear discriminant analysis
• Sparse linear discriminant analysis by thresholding for high dimensional data
• A tail inequality for quadratic forms of subgaussian random vectors
• Smooth discrimination analysis
• Measuring mass concentrations and estimating density contour clusters -- An excess mass approach
• A Road to Classification in High Dimensional Space: The Regularized Optimal Affine Discriminant
• Isotropic local laws for sample covariance and generalized Wigner matrices
• Neyman-Pearson classification, convexity and stochastic constraints
• A Neyman–Pearson Approach to Statistical Learning
• Variance and covariance inequalities for truncated joint normal distribution via monotone likelihood ratio and log-concavity
• A plug-in approach to Neyman-Pearson classification
• Analysis to Neyman-Pearson classification with convex loss function

Cited In (6)

• Neyman-Pearson classification under high-dimensional settings
• Asymmetric Error Control Under Imperfect Supervision: A Label-Noise-Adjusted Neyman–Pearson Umbrella Algorithm
• Title not available (Why is that?)
• A plug-in approach to Neyman-Pearson classification
• nproc
• Model-averaging-based semiparametric modeling for conditional quantile prediction

Uses Software

• R
• penalizedLDA

This page was built for publication: Neyman-Pearson classification: parametrics and sample size requirement

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4969044)