Road Traffic Incident Detection Model Based on SMO

Road traffic automatic incident detection system is the important composition of a road traffic monitoring system. On the basis of support vector machines (SVM) theory and sequential minimal optimization (SMO) algorithm, the road traffic incident detection model based on SMO was put forward. With Visual C++ and Matlab, a simulation experiment was applied to the model. The shortcomings of quadratic programming (QP) algorithm were analyzed. The effects of three different kernel functions on sample training SVM classifier and on detection performance were compared using Chunking algorithm and SMO algorithm. The results show that Gauss kernel function is better than the other two kernel functions for the performance of training and incident detection using SMO.

• Record URL:
  https://doi.org/10.1061/41039(345)130
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/isbn/9780784410394
• Supplemental Notes:
  • © 2009 American Society of Civil Engineers.
• Corporate Authors:

  American Society of Civil Engineers

  1801 Alexander Bell Drive
  Reston, VA  United States  20191-4400
• Authors:
  • Cong, Haozhe
  • Fang, Shouen
  • Guo, Jing
• Conference:
  • Second International Conference on Transportation Engineering
  • Location: Chengdu , China
  • Date: 2009-7-25 to 2009-7-27
• Publication Date: 2009-7

Language

• English

Media Info

• Media Type: Web
• Pagination: pp 782-787
• Monograph Title: International Conference on Transportation Engineering 2009

Subject/Index Terms

• TRT Terms: Algorithms; Automatic incident detection; Simulation
• Uncontrolled Terms: Kernel functions; Quadratic programming; Sequential minimal optimization; Support vector machines
• Subject Areas: Highways; Operations and Traffic Management; I73: Traffic Control;

Filing Info

• Accession Number: 01525235
• Record Type: Publication
• ISBN: 9780784410394
• Files: TRIS, ASCE
• Created Date: Nov 12 2013 1:36PM