Table 4 |
|||||
|
Best-fit ARIMA models and their predictors by patient acuity category |
|||||
|
MAPE (%) |
|||||
|
|
|||||
|
Patient acuity category |
Best-fit model |
No. of predictors |
Predictors (maximum lag correlation) |
Test |
Validation |
|
|
|||||
|
P1 |
ARIMA(0,1,1) |
1 |
PSI > 50 (2 days) |
18.2 |
16.8 |
|
P2 |
ARIMA(1,1,1)(1,0,1) |
1 |
Public holiday (1 day) |
7.7 |
6.7 |
|
P3 |
ARIMA(0,1,1)(1,0,1) |
2 |
Public holiday (1 day), PSI > 50 (0 day) |
7.2 |
8.6 |
|
|
|||||
|
All |
ARIMA(0,1,1)(1,0,1) |
1 |
Public holiday (1 day) |
4.4 |
4.8 |
|
|
|||||
|
ARIMA: auto-regression integrated moving average MAPE: mean absolute percentage error (p, d, q)(P, D, Q): p is the order of auto-regression, d is the order of differencing (integration), and q is the order of moving average; P, D, Q are their seasonal counterparts |
|||||
|
Sun et al. BMC Emergency Medicine 2009 9:1 doi:10.1186/1471-227X-9-1 |
|||||
