Mobility, Transportation and Navigation

[E4-01] Generating Human Daily Activity Sequences Using Neural Networks Models

*Weiying WANG¹, Toshihiro Osaragi¹ (1. Tokyo Institute of Technology)

キーワード：Human daily activities, Generation model, Neural networks, PT data

Human activity patterns have raised broad interests across disciplines, and they have been studied using various data sources in terms of different attributes. In the field of traffic research, it is widely believed that most travel behaviors are driven by activities to be conducted. In this paper, we present neural network models to generate synthetic datasets for daily activity sequences, which may be further utilized to replicate daily people’s movements in the city. Generation results are compared, and the models are discussed in detail.