1 Introduction
2 Data collection and extraction
3 Development of prediction scheme using SARIMA
3.1 Model identification
3.2 Model estimation and diagnostic checking
Model | Type | Parameters |
Z value | AIC | |
---|---|---|---|---|---|
(3,0,0) × (0,1,1) 144 | Non-seasonal AR | AR1
| 0.34 | 5.80 | 4218.7 |
AR2
| 0.27 | 4.59 | |||
AR3
| 0.07 | 1.27 | |||
Seasonal MA | MA1
| −0.41 | −3.38 | ||
(2,0,0) × (0,1,1) 144 | Non-seasonal AR | AR1
| 0.36 | 6.49 | 4218.3 |
AR2
| 0.30 | 5.36 | |||
Seasonal MA | MA1
| −0.40 | −3.37 | ||
(1,0,0) × (0,1,1) 144 | Non-seasonal AR | AR1
| 0.51 | 10.1 | 4243.6 |
Seasonal MA | MA1
| −0.33 | −3.13 |
4 Corroboration of the prediction scheme
Scenario | Number of previous days considered in the model to predict traffic flow on June 06, 2014 | ||||||||
---|---|---|---|---|---|---|---|---|---|
1 | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | ||||||
2 | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | |||||
3 | May 30 (Fri) | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | ||||
4 | May 29 (Thu) | May 30 (Fri) | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | |||
5 | May 28 (Wed) | May 29 (Thu) | May 30 (Fri) | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | ||
6 | May 27 (Tue) | May 28 (Wed) | May 29 (Thu) | May 30 (Fri) | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) | |
7 | May 26 (Mon) | May 27 (Tue) | May 28 (Wed) | May 29 (Thu) | May 30 (Fri) | June 02 (Mon) | June 03 (Tue) | June 04 (Wed) | June 05 (Thu) |