Back to Search Start Over

Time-series dynamic three-way group decision-making model and its application in TCM efficacy evaluation.

Authors :
Chu, Xiaoli
Sun, Bingzhen
Mo, Xiumei
Liu, Junfeng
Zhang, Yu
Weng, Heng
Chen, Dacan
Source :
Artificial Intelligence Review; Oct2023, Vol. 56 Issue 10, p11095-11121, 27p
Publication Year :
2023

Abstract

Clinical curative effect is the core value and fundamental pursuit of medicine. The data-driven clinical quantitative evaluation model is an important research issue in clinical efficacy evaluation. The existing clinical efficacy evaluation methods have made a lot of contributions to the structured information of static attributes, but less attention is paid to the heterogeneous information with dynamic and random characteristics. This paper discusses a class of clinical efficacy evaluation decision-making problems that contain both time-series and dynamic characteristics and uses the three-way decision principle to describe these characteristics. The group decision-making theory is introduced to aggregate the evaluation information of time series dynamic features generated in the decision-making process. A time series dynamic three-way group decision model for clinical efficacy evaluation is constructed, the optimal decision rules with minimum expected loss objective are calculated and proofed in the proposed model, and then the optimal treatment alternative is obtained. Finally, taking the decision-making problem of clinical efficacy evaluation of traditional Chinese medicine (TCM) treatment of atopic dermatitis as the research object, based on the actual clinical data of Guangdong Provincial Hospital of TCM, the decision-making process and steps of the theoretical model constructed in this paper are verified and applied. The data-driven quantitative clinical evaluation model has achieved better application effects than the actual clinical scenarios. The research results in this paper provide new ideas and theoretical methods for real-world clinical efficacy evaluation research. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
02692821
Volume :
56
Issue :
10
Database :
Complementary Index
Journal :
Artificial Intelligence Review
Publication Type :
Academic Journal
Accession number :
170039900
Full Text :
https://doi.org/10.1007/s10462-023-10445-z