코플라 함수를 이용한 이차원 보증 정책 모형 연구

Purpose: This study aims to propose a novel method for constructing a warranty cost analysis model using copula functions. It explores the properties of selected copula models and applies them to fit two-dimensional warranty data, thereby enhancing the modeling of dependent failure behaviors. Method...

Full description

Saved in:
Bibliographic Details
Published in품질경영학회지, 53(3) pp. 287 - 297
Main Authors 박민재, 김승범
Format Journal Article
LanguageKorean
Published 한국품질경영학회 01.09.2025
Subjects
Online AccessGet full text
ISSN1229-1889
2287-9005
DOI10.7469/JKSQM.2025.53.3.287

Cover

More Information
Summary:Purpose: This study aims to propose a novel method for constructing a warranty cost analysis model using copula functions. It explores the properties of selected copula models and applies them to fit two-dimensional warranty data, thereby enhancing the modeling of dependent failure behaviors. Methods: Copulas provide a flexible framework for modeling dependency structures among random variables with different marginal distributions. By employing copula functions, a joint probability distribution was formulated, which served as the foundation for developing a two-dimensional warranty cost model. Results: The proposed approach significantly improves the prediction accuracy of bivariate warranty claims by capturing complex inter-variable dependencies that traditional models fail to represent. This enhanced modeling capability allows for more reliable cost estimation and better-informed warranty management decisions. Conclusion: This study introduces a new methodology for constructing copula-based models tailored to two-dimensional reliability and warranty data. The proposed framework was validated using real warranty claim data from an automotive dealer, demonstrating its practical applicability and potential value in both academic and industrial settings. KCI Citation Count: 0
Bibliography:http://jksqm.org/journal/view.php?doi=10.7469/JKSQM.2025.53.3.287
ISSN:1229-1889
2287-9005
DOI:10.7469/JKSQM.2025.53.3.287