With the fast development trend of highly integrated electronic products, as the key technology of three-dimensional(3D) interconnect circuits, the research on Monolithic Inter-tier Via(MIV) testing technology is particularly critical. Therefore,this paper proposes a machine-learning-based MIV open fault localization method. This method adopts the EXtreme Gradient Boosting (XGBoost) prediction model optimized based on the Grid Search (GS) algorithm, and takes the S-parameters as the training dataset. The S-parameters are used as the training data set to train and evaluate the prediction model. This method can effectively solve the problem that the traditional MIV test method is difficult to accurately locate the MIV open faults. Simulation results show that the prediction model proposed in this paper can accurately predict the location of open faults.