城市活力嵌入及其在交通预测中的应用

Urban vibrancy reflects dynamic human activities within urban spaces and is often measured using mobile data that captures trends in floating populations. This study proposes a novel approach to derive Urban Vibrancy embeddings from real-time floating population data to enhance traffic prediction models. Specifically, variational autoencoders (VAE) are utilized to compress this data into actionable embeddings, which are then integrated with long short-term memory (LSTM) networks to predict future embeddings. These embeddings are subsequently applied in a sequence-to-sequence framework for traffic forecasting. The contributions of this research are threefold: the use of principal component analysis (PCA) to interpret the embeddings, revealing temporal patterns such as variations in floating populations on weekdays and weekends, thus significantly improving the accuracy of traffic predictions.