CONTACT US: Contact info
|
Citation |
Khasnabis S, Lyoo SH. Transp. Res. Rec. 1989; 1249: 30-36. |
|
Copyright |
(Copyright © 1989, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing) |
|
DOI |
unavailable |
PMID |
unavailable |
|
Abstract |
The purpose of this paper is to test the feasibility of using the Box-Jenkins method of time series analysis for forecasting truck accidents. Time series analysis is a technique by which the autocorrelation between sequential observations is analyzed and models are developed to produce forecasts. The authors used the Auto-Regressive Integrated Moving Average Method (ARIMA) in an effort to incorporate seasonal fluctuations in the data base to develop the model. A total of 88 data points representing monthly accidents involving large trucks in Michigan, observed between January 1978 and April 1985, was used to develop the model. Two types of checks were used to test the goodness of fit of the model. First, diagnostic checks were conducted to test the degree of correspondence between the observed data (used for model development) and the model output. This test indicated that approximately 70% of the autocorrelations are accounted for in the model. Second, the model was used to forecast monthly accident data for the 20-month period between May 1985 and December 1986 (data base not used in model development). The forecast data were then compared with the actual truck accidents observed during the same period. This test showed excellent correspondence between the observed data and the model output. The authors recommend further studies to test the feasibility of using time series analysis as an accident prediction tool. Language: en |
|
Keywords |
Mathematical Models; Accidents - Prediction; Motor Trucks; Statistical Methods - Time Series Analysis |