vix.ing · top · new · best · stats · spec

Stability Properties of Feature Selection Measures

2024/05/01 by Alexander Bulinski, A. V. Bulinski
Computer Science · Engineering · Mathematics · #Bayesian Methods and Mixture Models #Control Systems and Identification #Statistical Methods and Inference

paper · doi:10.1137/s0040585x97t991726

crossref issued 2024/05/01 · crossref published 2024/05/01 · crossref published-print 2024/05/01 · openalex publication_date 2024/05/01 · crossref published-online 2024/05/02 · crossref created 2024/05/02 · crossref deposited 2024/05/02 · openalex created_date 2025/10/10 · crossref indexed 2026/07/27 · openalex updated_date 2026/07/27

Abstract

In this paper, we prove that the monotonicity property of the stability measure for the feature (factor) selection introduced by Nogueira, Sechidis, and Brown [J. Mach. Learn. Res., 18 (2018), pp. 1--54] may not hold. Another monotonicity property takes place. We also show the cases in which it is possible to compare by certain parameters the matrices describing the operation of algorithms for identifying relevant features.

Citations