Fitting Time Series Models to Fisheries Data to Ascertain Age
Table 2
Model diagnostics for the ARIMA (p, 1, q) model.
p\q
0
1
2
3
(a)
0
−80.6663
−83.841
−81.8519
−88.0023
1
−83.211
−81.8484
−90.054
−89.9316
2
−82.1529
−96.3694
−83.3204
−117.292
3
−80.8888
−99.4612
−107.115
−104.951
(b)
0
0.017768
0.016501
0.016498
0.014002
1
0.016651
0.016499
0.013748
0.013359
2
0.016426
0.012415
0.014449
0.008054
3
0.016247
0.011402
0.010011
0.010014
(c)
0
42.33315
44.92048
44.92594
49.00115
1
44.60548
44.92422
50.02699
50.96579
2
45.07647
53.18471
47.66019
65.64576
3
45.44441
55.73059
60.5575
60.47542
(d)
0
−84.66630
−85.56429
−81.29855
−85.17230
1
−84.93434
−81.29511
−87.22398
−84.82492
2
−81.59961
−93.53942
−78.21372
−109.90819
3
−78.05882
−94.35452
−99.73167
−95.29084
The bold values show the best model parameters, while the italic values show the second best model parameters. (a) The AICs for selected ARIMA models. (b) The MSE for selected ARIMA models. (c) The log likelihood for selected ARIMA models. (d) The BIC for selected ARIMA models.