Prediction of bus travel times is of crucial importance for passengers in letting them know their departure time from an origin and arrival time at a destination and allowing them to make decisions (e.g., postpone departure time at certain hours) and to reduce their waiting time at bus stops. To predict bus travel times, it is important to know whether the target routes are stable or not. In this paper, we propose a time series approach to predict the travel time over an interval between two adjacent bus stops. We build Artificial Neural Network (ANN) models to predict the travel time over the interval. To make accurate predictions, we divide a day into 8 time-periods in calculating travel time over the interval and classify unstable intervals into three types: weak, medium and strong unstable. We use bus probe data collected from November 21st to December 20th 2013 and provided by Nishitetsu Bus Company, Fukuoka, Japan. Experimental results show that our models can effectively improve the prediction accuracy of travel times over intervals by focusing on the three unstable classes and calculating travel times for each interval at each of 8 time-periods in a day.