Generative Transformer Model Synthesizes Human Mobility and Daily Activity Chains

A generative Transformer model uses socio-demographic and household attributes to synthesize complete daily trajectories. The architecture combines a semantic activity chain module with a spatial location module to transfer learned national patterns to specific regions like California and Mexico.