Research Article
A Fast and Efficient Estimation of the Parameters of a Model of Accident Frequencies via an MM Algorithm
Table 3
Results for Scenario
3 (
and
). Values in brackets are standard deviations.
| | MM | SqS3 | NR | BFGS |
| | Convergence proportion (%) | 100 | 100 | 99.7 | 28 | | Iterations | 30.2 (3.9) | 7.9 (0.7) | 8.6 (2.5) | 14 (0.2) | | CPU time (secs) | 0.009 | 0.004 | 0.043 | 0.426 | | Time ratio | 1.00 | 0.50 | 4.90 | 48.60 | | Log-likelihood | -831.26 | -831.26 | -831.27 | -831.07 | | MSE | | | | |
| | Convergence proportion (%) | 100 | 100 | 99.6 | 30.9 | | Iterations | 39.5 (3.6) | 9.2 (0.8) | 8.8 (2.3) | 16 (6.4) | | CPU time (secs) | 0.012 | 0.005 | 0.045 | 0.669 | | Time ratio | 1.00 | 0.41 | 3.70 | 55.22 | | Log-likelihood | -84027.58 | -84027.58 | -84027.15 | -88978.38 | | MSE | | | | |
|
|
