The prediction model for Preterm Birth(PTB) and Low Birth Weight(LBW) infants has been built by adopting Machine Learning(ML) techniques including Logical Regression(LR), Support Vector Machine(SVM) and Random Forest(RF) algorithm. The paper figures out the optimal models of different algorithms through using cross validation and then evaluates the predication performance of 3 models according to accuracy rate, F1score and AUC value. The result shows that the model based on RF algorithm features the best prediction effect.