A diabetes prediction model is constructed based on feature selection algorithm and artificial neural network, and the steps and methods of model construction and evaluation are expounded. Sensitivity, specificity, accuracy and ROC-AUC are used as evaluation metrics to evaluate the performance of the proposed model, and the proposed model is compared with other algorithms through experiments. The result indicates that the diabetes prediction model based on feature selection and artificial neural network has better anti-interference capability and predictive performance, which is more suitable for complex datasets with unknown influencing indicators.