TY - JFULL AU - Muhammad Awais Shafique and Eiji Hato and Hideki Yaginuma PY - 2014/11/ TI - Using Probe Person Data for Travel Mode Detection T2 - International Journal of Computer and Information Engineering SP - 1777 EP - 1782 VL - 8 SN - 1307-6892 UR - https://publications.waset.org/pdf/9999446 PU - World Academy of Science, Engineering and Technology NX - Open Science Index 94, 2014 N2 - Recently GPS data is used in a lot of studies to automatically reconstruct travel patterns for trip survey. The aim is to minimize the use of questionnaire surveys and travel diaries so as to reduce their negative effects. In this paper data acquired from GPS and accelerometer embedded in smart phones is utilized to predict the mode of transportation used by the phone carrier. For prediction, Support Vector Machine (SVM) and Adaptive boosting (AdaBoost) are employed. Moreover a unique method to improve the prediction results from these algorithms is also proposed. Results suggest that the prediction accuracy of AdaBoost after improvement is relatively better than the rest. ER -