The powerful performance of deep learning is evident to all. With the deepening of research, neural networks have become more complex and not easily generalized to resource-constrained devices. The emergence of a series of model compression algorithms makes artificial intelligence on edge possible. Among them, structured model pruning is widely utilized because of its versatility. Structured pruning prunes the neural network itself and discards some relatively unimportant structures to compress the model’s size. However, in the previous pruning work, problems such as evaluation errors of networks, empirical determination of pruning rate, and low retraining efficiency remain. Therefore, we propose an accurate, objective, and efficient pruning algorithm—Combine-Net, introducing Adaptive BN to eliminate evaluation errors, the Kneedle algorithm to determine the pruning rate objectively, and knowledge distillation to improve the efficiency of retraining. Results show that, without precision loss, Combine-Net achieves 95% parameter compression and 83% computation compression on VGG16 on CIFAR10, 71% of parameter compression and 41% computation compression on ResNet50 on CIFAR100. Experiments on different datasets and models have proved that Combine-Net can efficiently compress the neural network’s parameters and computation.

With the rapid advancement of technology, modeling travel behavior based on large-scale mobile trajectory data has become a core issue in urban traffic management. Mining complex patterns of travel behavior and modeling based on this to explain, reproduce, and predict human mobility is of great significance. Interpretability and universality are the goals in modeling the complexity of travel behavior. Although advanced deep learning models are increasingly popular in human mobility prediction, the complex model structures, high training costs, and the trade-off between accuracy and generalization are unavoidable challenges. This paper proposes a straightforward, interpretable large-scale human mobility prediction model. By modeling periodic pattern time attenuation and spatiotemporal correlations of destination points, this model enables efficient and general real-time prediction of massive individual complex trajectories. The proposed method was applied to trajectory data from an anonymous metropolitan area in Japan. Comparative analysis with mainstream deep learning models revealed the robust and outstanding performance of our method across local and global similarity evaluation metrics. This approach is expected to serve as a universal modeling strategy for the complex trajectories of massive individuals to explore human mobility patterns.