Fuzzy k-nearest neighbors with monotonicity constraints: Moving towards the robustness of monotonic noise

NEUROCOMPUTING

卷:439

Article

Gonzalez, Sergio¹  Garcia, Salvador¹  Li, Sheng-Tun^2,3,4  John, Robert⁵  Herrera, Francisco^1,6 

[1] Univ Granada, Dept Comp Sci & Artificial Intelligence, Granada 18071, Spain

[2] Natl Cheng Kung Univ, Dept Ind & Informat Management, Tainan 701, Taiwan

[3] Natl Cheng Kung Univ, Inst Informat Management, Tainan 701, Taiwan

[4] Natl Cheng Kung Univ, Ctr Innovat FinTech Business Models, Tainan 701, Taiwan

[5] Univ Nottingham, Sch Comp Sci, ASAP Res Grp, Nottingham NG8 1BB, England

[6] King Abdulaziz Univ, Fac Comp & Informat Technol, Jeddah, Saudi Arabia

关键词: Fuzzy k-NN;    Monotonic constraints;    Ordinal classification;    Ordinal regression;    Class noise;   

DOI  :  10.1016/j.neucom.2019.12.152

来源: Elsevier

【 摘 要 】

This paper proposes a new model based on Fuzzy k-Nearest Neighbors for classification with monotonic constraints, Monotonic Fuzzy k-NN (MonFkNN). Real-life data-sets often do not comply with monotonic constraints due to class noise. MonFkNN incorporates a new calculation of fuzzy memberships, which increases robustness against monotonic noise without the need for relabeling. Our proposal has been designed to be adaptable to the different needs of the problem being tackled. In several experimental studies, we show significant improvements in accuracy while matching the best degree of monotonicity obtained by comparable methods. We also show that MonFkNN empirically achieves improved perfor-mance compared with Monotonic k-NN in the presence of large amounts of class noise.& nbsp; (c) 2021 Elsevier B.V. All rights reserved.

【 授权许可】

Free