In order to solve the problem of the long monitoring and measurement cycle of surface settlement in the shallow-buried section of expressway tunnels, the convolutional neural network algorithm was introduced into the research on the prediction of surface settlement values in the shallow-buried section of expressways. Based on the measured data of surface settlement in the shallow-buried section of the Lanyaping tunnel of the Kaizhou-Yunyang expressway project, 70% of the measured data was randomly selected as the training set, and 30% of the measured data was used as the test set. The engineering example shows that when the convolutional neural network algorithm is applied to the research on the prediction of surface settlement values in the shallow-buried section of expressways, the predicted values of surface settlement in the shallow-buried section of the tunnel can be calculated quickly. Moreover, the prediction results indicate that the root mean square error of the prediction for the training set is 0.0436 88, and that for the test set is 0.173 38, which proves the applicability of the convolutional neural network algorithm.