摘要Objective:To explore factors influencing the Qi-stagnation constitution in traditional Chinese medicine(TCM)through mathematical and computational methods,aiming to early prevention and treatment of related diseases.Methods:The traditional grey correlation analysis,improved with the entropy weight method and Euclidean distance,was applied to investigate factors influencing the Qi-stagnation constitution.The top five factors identified before and after the improvement were used to construct random forest models.Results:A total of 1342 Qi-stagnation constitution identification data sets were collected.The grey correlation degree of each influencing factor progressively decreased during the transition from the traditional to the improved grey correlation analysis method.The random forest model built using the traditional grey correlation analysis method achieved prediction accuracies of 75.61%for the training set and 67.99%on the validation set.The random forest model developed using the improved grey correlation analysis method achieved prediction accuracies of 89.78%on the training set and 81.64%on the validation set,outperforming the model based on the traditional methods.Conclusions:The improved grey correlation analysis method is more suitable for analyzing the influencing factors of the Qi-stagnation constitution.This study offers a new approach and is a reference for research on TCM constitutions.