Method for acquiring target model, method for determining prognosis evaluation value, apparatuses, device, and medium

Abstract

Provided is a method for acquiring a target model, including: acquiring sample groups corresponding to sample users, the sample group including sample information of multiple modalities; performing iterative trainings on a preset model based on the sample groups to obtain the target model, the target model being configured to predict a prognosis evaluation value of a target object, in each iterative training, feature extraction is performed on the sample information of the multiple modalities in a current sample group using the preset model, a predicted prognosis evaluation value and a consistency expression value are determined based on extracted sample features, the consistency expression value is used for characterizing a consistency degree of the sample features corresponding to a same target disease; and parameters of the preset model are updated based on the predicted prognosis evaluation value, a prognosis evaluation label corresponding to the current sample group, the consistency expression value.