Detailed Information

Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

이변수 함수에 대한 최소 절대 편차 퍼지 변환

Full metadata record
DC Field Value Language
dc.contributor.author민희준-
dc.contributor.author김성아-
dc.contributor.author정혜영-
dc.date.accessioned2023-07-05T05:41:04Z-
dc.date.available2023-07-05T05:41:04Z-
dc.date.issued2022-12-
dc.identifier.issn1976-9172-
dc.identifier.issn2288-2324-
dc.identifier.urihttps://scholarworks.bwise.kr/erica/handle/2021.sw.erica/113178-
dc.description.abstract본 논문에서는 퍼지 변환과 최소 절대 편차 방법을 결합한 최소 절대 편차 퍼지 변환의 목적함수를 최적화하기 위한 알고리즘을 제시한다. 퍼지 변환은 데이터를 축소 및 복원하는방법이며 퍼지 변환과 최소 제곱법을 결합한 최소 제곱 퍼지 변환은 이상치에 강건하지 않다는 단점을 가지고 있다. 이를 해결한 기법으로 이상치에 강건한 최소 절대 편차 퍼지 변환이 소개되었다. 본 논문은 이변수 함수에 대한 최소 절대 편차 퍼지 변환 알고리즘을 구현하였으며 실험을 통해 최소 절대 편차 퍼지 변환이 최소 제곱 퍼지 변환보다 이상치에 강건함을 검증하였다. 또한, 세 가지 퍼지 변환 기법의 데이터 복원력에 대해 비교하였다.-
dc.description.abstractIn this paper, we present an algorithm to optimize the objective function of theLeast Absolute Deviation Fuzzy Transform, which applied least absolute deviationapproximation method to Fuzzy Transform. Fuzzy Transform is a method of datareduction and reconstruction, Least Squares Fuzzy Transform which applied leastsquares method to fuzzy transform has the disadvantage of not being robust tooutlier. Least Absolute Deviation Fuzzy Transform, which is robust to outlier hasbeen proposed to overcome the disadvantage. We present the algorithm of LeastAbsolute Deviation Fuzzy Transform and experiments show that the Least AbsoluteDeviation Fuzzy Transform is robust to outlier than Least Squares FuzzyTransform. In addition, we compare the reconstruction performance of three fuzzytransform based methods. augmentation methods are lack of the consistency of generated data due to theunstability of model convergence. This paper proposes a novel method thateffectively augments labeled time-series data by a channeling idea, whichaugments the original data without deformation, ensuring the consistency of thedata. Experimental results showed that the proposed method can detect processrisks about 8.1 to 16.7% more accurately than comparative models.-
dc.format.extent6-
dc.language한국어-
dc.language.isoKOR-
dc.publisher한국지능시스템학회-
dc.title이변수 함수에 대한 최소 절대 편차 퍼지 변환-
dc.title.alternativeLeast Absolute Deviation Fuzzy Transform for functions of two variables-
dc.typeArticle-
dc.publisher.location대한민국-
dc.identifier.doi10.5391/JKIIS.2022.32.6.530-
dc.identifier.bibliographicCitation한국지능시스템학회 논문지, v.32, no.6, pp 530 - 535-
dc.citation.title한국지능시스템학회 논문지-
dc.citation.volume32-
dc.citation.number6-
dc.citation.startPage530-
dc.citation.endPage535-
dc.identifier.kciidART002906038-
dc.description.isOpenAccessY-
dc.description.journalRegisteredClasskci-
dc.subject.keywordAuthor퍼지 변환-
dc.subject.keywordAuthor최소 제곱 퍼지 변환-
dc.subject.keywordAuthor최소 절대 편차 퍼지 변환-
dc.subject.keywordAuthor최소 절대 편차 방법-
dc.subject.keywordAuthor이상치-
dc.subject.keywordAuthorFuzzy Transform-
dc.subject.keywordAuthorLeast Squares Fuzzy Transform-
dc.subject.keywordAuthorLeast Absolute Deviation Fuzzy Transform-
dc.subject.keywordAuthorLeast Absolute Deviation Approximation-
dc.subject.keywordAuthorOutlier-
dc.identifier.urlhttps://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE11179130-
Files in This Item
Go to Link
Appears in
Collections
COLLEGE OF SCIENCE AND CONVERGENCE TECHNOLOGY > ERICA 수리데이터사이언스학과 > 1. Journal Articles

qrcode

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.

Related Researcher

Researcher JUNG, HYE YOUNG photo

JUNG, HYE YOUNG
ERICA 소프트웨어융합대학 (ERICA 수리데이터사이언스학과)
Read more

Altmetrics

Total Views & Downloads

BROWSE