RUS  ENG
Full version
JOURNALS // Diskretnyi Analiz i Issledovanie Operatsii // Archive

Diskretn. Anal. Issled. Oper., 2021 Volume 28, Issue 2, Pages 60–73 (Mi da1277)

NP-hardness of some data cleaning problem

O. A. Kutnenkoab, A. V. Plyasunovab

a Sobolev Institute of Mathematics, 4 Acad. Koptyug Avenue, 630090 Novosibirsk, Russia
b Novosibirsk State University, 2 Pirogov Street, 630090 Novosibirsk, Russia

Abstract: We prove the NP-hardness of the problem of outliers detection considered in this paper, to solving which a data analysis problem is reduced. As a quantitative assessment of the compactness of the image, the function of rival similarity (FRiS-function) is used, which evaluates the local similarity of objects with their closest neighbors. Illustr. 1, bibliogr. 23.

Keywords: NP-hardness, detecting outliers, image compactness, function of rival similarity.

UDC: 519.87+519.854

Received: 10.06.2020
Revised: 22.12.2020
Accepted: 24.12.2020

DOI: 10.33048/daio.2021.28.692


 English version:
Journal of Applied and Industrial Mathematics, 2021, 15:2, 285–291

Bibliographic databases:


© Steklov Math. Inst. of RAS, 2025