One important component of an advanced traveler information system (ATIS) is the bus-arrival-time
information service. Dynamic at-stop real-time information plays an important role in waiting time reduction,
positive psychological factors, increased willingness to pay, travel behavior adjustment, mode choice, higher
customer satisfaction, and better brand effect (Dziekan & Kottenhoff, 2007).
As a basic component of bus-arrival-time, the accurate prediction/estimation of bus travel time has a
significant effect on the overall performance of the prediction results of bus-arrival-time. However, bus travel
time is affected by many factors, as validated in previous research. Generally speaking, those factors can be
clustered into five categories: passenger factors, such as boarding, alighting time (Tétreault & A.El-Geneidy,