TY - JOUR
T1 - A phase-II change-point-based distribution-free scheme for monitoring of three process aspects
AU - Chen, Xinran
AU - Mukherjee, Amitava
AU - Xiang, Dongdong
AU - Li, Wendong
N1 - Publisher Copyright:
© 2024 American Statistical Association and Taylor & Francis.
PY - 2025
Y1 - 2025
N2 - In recent years, more attention has been paid to joint monitoring of two-aspects, namely, location and scale parameters, of a continuous process when their in-control (IC) standard values are unknown. However, barring a few, most literature ignores the third aspect of the process, namely the shape parameter. This paper proposes a new distribution-free change point formulation-based scheme for monitoring three aspects of a process simultaneously. No prior knowledge about the process distribution or its parameters, except for continuousness, is required. We discuss the in-control robustness of the proposed scheme, assuming that an independently and identically distributed Phase-I sample of size greater or equal to five from an unknown population is available. Simulation studies based on Monte-Carlo reveal some advantages of the proposed scheme over its competitors. An illustration with the arrival delay data of a Railway Network is presented, and some concluding remarks with future research directions are offered.
AB - In recent years, more attention has been paid to joint monitoring of two-aspects, namely, location and scale parameters, of a continuous process when their in-control (IC) standard values are unknown. However, barring a few, most literature ignores the third aspect of the process, namely the shape parameter. This paper proposes a new distribution-free change point formulation-based scheme for monitoring three aspects of a process simultaneously. No prior knowledge about the process distribution or its parameters, except for continuousness, is required. We discuss the in-control robustness of the proposed scheme, assuming that an independently and identically distributed Phase-I sample of size greater or equal to five from an unknown population is available. Simulation studies based on Monte-Carlo reveal some advantages of the proposed scheme over its competitors. An illustration with the arrival delay data of a Railway Network is presented, and some concluding remarks with future research directions are offered.
KW - Change-point detection
KW - distribution-free
KW - linear rank statistic
KW - statistical process control
KW - tri-aspect
UR - https://www.scopus.com/pages/publications/86000376865
U2 - 10.1080/10485252.2024.2365212
DO - 10.1080/10485252.2024.2365212
M3 - 文章
AN - SCOPUS:86000376865
SN - 1048-5252
VL - 37
SP - 128
EP - 147
JO - Journal of Nonparametric Statistics
JF - Journal of Nonparametric Statistics
IS - 1
ER -