Double Articulation Analysis and Prediction of Driving Behavior

The authors describe in this paper the recent results of experiments using the double articulation analyzer to analyze and predict the driving behavior. Upon the observations by previous researches showing that the prediction of actual value of driving behavior is difficult, the authors focus on predicting discrete symbols of driving behavior inferred by a probabilistic model. The authors' method incorporates longer temporal context using unsupervised learning algorithm of hierarchical n-gram model which was developed in Natural Language Processing field and extends it to the prediction method. The experimental results show that the prediction based on the assumption of the double articulation structure in driving behavior data is effective for the real driving data.

  • Availability:
  • Supplemental Notes:
    • Abstract used with permission of ITS Japan. Paper No. 4170.
  • Corporate Authors:

    ITS Japan

    Tokyo,   Japan 
  • Authors:
    • Hitomi, Kentarou
    • Bando, Takashi
    • Takenaka, Kenro
  • Conference:
  • Publication Date: 2013


  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References;
  • Pagination: 10p
  • Monograph Title: 20th ITS World Congress, Tokyo 2013. Proceedings

Subject/Index Terms

Filing Info

  • Accession Number: 01538851
  • Record Type: Publication
  • ISBN: 9784990493981
  • Files: TRIS
  • Created Date: Sep 26 2014 2:24PM