Method and computing device of establishing model for analyzing correlation of white coat effect, and method and computing device of analyzing correlation of white coat effect capable of conveniently and efficiently predicting probability of tester experiencing white coat effect

Provided is a method of establishing a model for analyzing the correlation of the white coat effect. For each of a plurality of testers, the physiological data, medication data and labeling results corresponding to the tester are used as a set of training data. Then, based on the training data, an interpretable machine learning algorithm is used to obtain a correlation model and a plurality of influence levels. Next, from each set of training data, the item with the lowest impact is removed, and a machine learning algorithm and a verification method are used to obtain a white coat effect correlation model and a model evaluation value based on the training data after removal. Furthermore, the item corresponding to the lowest impact is removed, and another white coat effect correlation model and another model evaluation value are obtained. Finally, a target white coat effect correlation model is obtained based on these model evaluation values.