Time Series Model Package
TSMOUTOPT Object
The TSMOUTOPT object collects the optimization history, which contains training and validation errors, from a TSM object, TSMObj, and stores them in an output table. This information is useful for evaluating how well the model fits the dependent series. This object is available only for the RNNSPEC object.
Table 21 shows the contents of the TSMOUTOPT object.
Table 21: Contents of the TSMOUTOPT Object
| Column | Type | Description |
|---|---|---|
| Epoch | Numeric | Epoch number |
| Train Error | Numeric | Training error |
| Valid Error | Numeric | Validation error |
Table 22 summarizes the methods that are associated with the TSMOUTOPT object.
Table 22: Methods of the TSMOUTOPT Object
| Method | Description |
|---|---|
| Collect | Collects the optimization history from the TSM object |
Figure 13 illustrates the data flow through the TSMOUTOPT object.
Figure 13: TSMOUTOPT Object Data Flow
