Research Article

Application of the Empirical Bayes Method with the Finite Mixture Model for Identifying Accident-Prone Spots

Table 2

Modeling results of NB models for Periods 1 and 2.

EstimatesPeriod 1 (years 1997 and 1998) Period 2 (years 1999, 2000, and 2001)
ValueSEValueSE

Intercept −7.8360.497−8.1160.453
Ln(average daily traffic) 1.0930.0541.1370.048
Lane width −0.0440.020−0.0550.018
Total shoulder width −0.0130.004−0.0120.004
Curve density 0.0260.0140.0240.013
0.8250.0620.7920.049
Log-likelihood2924.4903409.444
AIC5860.9806830.880
BIC5892.8506862.763

SE: standard error.