关键词：汽车；大规模数据集；GPS
摘 要：The goal of this proposal is to develop novel modeling techniques to infer individual activity patterns from the large scale cell phone datasets and taxi data from NYC. As such this research offers a paradigm shift from traditional transportation modeling by using large scale, disaggregate data and provides an unique perspective to understand the complex interactions among human behavior, urban environments and traffic patterns. Urban development shapes the transportation systems, it determines what kind of transportation system a city has, and what does it look like. As an important dynamic component in urban systems, activities of transportation systems in turn captures the dynamics of the entire urban systems and enhance of our knowledge about the complex urban systems. This will ultimately contribute to the improvement of level of service and policy making on transportation systems. Taxi as a transportation tool has its unique characteristics. It is capable of capturing urban movement patterns both spatially and temporally since they serve as real time probes in the network. Moreover, we are able to examine the pulse of the city, the gap between supply and demand, real time road congestions and even more. On the other hand, accurate estimation and prediction of urban link travel times are important for improving urban traffic operations and identifying key bottlenecks in the traffic network.