Saturday, 30 January 2010

Wseas Transactions

New Subscription to Wseas Transactions

The following information was submitted:

Transactions: WSEAS TRANSACTIONS ON INFORMATION SCIENCE AND APPLICATIONS
Transactions ID Number: 89-385
Full Name: Omer Faruk Cansiz
Position: Assistant Professor
Age: ON
Sex: Male
Address: Department of Civil Engineering, Faculty of Engineering, Mustafa Kemal University., Iskenderun, 31200 Hatay, Turkey.
Country: TURKEY
Tel: 326 613 56 00 - 42 25
Tel prefix: +90
Fax:
E-mail address: ofcansiz@mku.edu.tr
Other E-mails: ofcansiz@gmail.com
Title of the Paper: New Inquiry on Prediction Model of Fatality Number in Traffic Accidents
Authors as they appear in the Paper: Omer Faruk Cansiz
Email addresses of all the authors: ofcansiz@mku.edu.tr
Number of paper pages: 15
Abstract: The Smeed Equation (SE) is the first model improved to estimate the dead number in accidents which consists of the independent variables of population and number of vehicles and the dependent variable of dead number. In this study, population variable in SE is replaced with vehicle-kilometer (vehicle-km). At first, SE is made suitable for the USA data and Revised SE is obtained. Then the coefficients are calculated again by the replacement of the vehicle-km with population and Improved SE is obtained. Afterwards, Artificial Neural Network (ANN) models are formed in both variable groups of population and vehicle-km. The best ANN model, whose inputs are number of vehicles and vehicle-km, has 15 neuron, tan-sig and purelin transfer functions and LM train algorithm. In the comparison of the best ANN model and Improved SE models, the value of R2 increases from 0,9428 to 0,9793, the value of mean square errors (MSE) decreases from 36.414 to 27.337 and the value of averag!
e percent errors (APE) decreases from 20,31% to 19,71%. As a result the replacement of the vehicle-km variable with population has a contribution in the estimation of the dead numbers in traffic accidents. This study showed that use of vehicle-km instead of the population in the dead number prediction can be improved accuracy of the proposed models. Moreover, ANN models can be used to predict the dead number in traffic accident with high correlation coefficient, low MSE and APE according to the SE and Loglinear regression methods.
Keywords: Motor-vehicle accident; Fatality in traffic accidents; Artificial neural network, Vehicle-km
EXTENSION of the file: .doc
Special (Invited) Session: Use of Artificial Neural Network to Estimate Number of Persons Fatally Injured
Organizer of the Session: 101-151
How Did you learn about congress: 1.Assoc.Prof. Ozgur Kisi -kisi@erciyes.edu.tr,2.Asst.Prof. Erdoðan Özbay-ozbay@gantep.edu.tr,3.Assoc.Prof. Ahmet Öztaþ-aoztas@gantep.edu.tr
IP ADDRESS: 88.249.53.41