Optional parameters for the ATT forecast engine
This section describes the optional parameters to configure the ATT forecast engine.
Use these parameters to modify the default values of the forecast
engine, when forecast monitoring is required:
| Display Name | Used by (Algorithms) | Accepted range | Engine Default | Description |
|---|---|---|---|---|
| Least Squares outlier standard deviation | BEST, LS | integer | 0 |
ModelDataClass.LeastSquaresOutlierSD_: The
minimum number of standard deviations for a data point when compared to the
forecast, to flag the same as an outlier for the Least Squares algorithm
(default = 0, which effectively stops the outlier detection). If the outlier
testing is required, low values are recommended, for example, 2. Outliers work
only when large errors occur. This is due to the Least Squares model fit
including any outlier observations in the initial calculation and therefore
being relatively unresponsive to outliers.
|
| Moving Average outlier standard deviation | BEST, MA, EWMA | integer | 0 |
ModelDataClass.movingAverageOutlierSD_: The
minimum number of standard deviations for a data point when compared to the
forecast, to flag the same as an outlier for the Moving Average algorithm
(default = 0, which effectively stops outlier detection). If outlier testing is
required, low values are recommended, for example, 2. Outliers work only when
large errors must occur. This is due to the Moving Average model fit including
any outlier observations in the initial calculation and therefore being
relatively unresponsive to outliers.
|
| Perform Outlier Testing | BEST, MA, EWMA, LS | true or false | true |
MainDataClass.performOutlierTesting_:
Determines if checks are performed for outliers in the Holt-Winters, Least
Squares, Moving Average, and Exponential Moving Average algorithms.
|
| Perform Step Changes | BEST, H-W | true or false | true |
MainDataClass.performStepChanges_: Determines
if checks are performed for step changes in the Holt-Winters algorithm.
|
| Perform Tracking | BEST, H-W, MA, EWMA, LS | true or false | yes |
MainDataClass.performTracking_: Determines if
the engine checks for tracking signals in the Holt-Winters, Least Squares,
Moving Average, and Exponential Moving Average algorithms.
|
| Step Change Maximum Run Length | BEST, H-W | integer | 3 |
ModelDataClass. stepChangeMaximumRunLength_:
Contains the maximum number of contiguous outliers required to trigger a step
change in the Holt-Winters algorithm. All the outliers must be in the same
direction.
|
| Tracking Type | BEST, H-W, MA, EWMA, LS | NOTRACKING, SIMPLETRACKINGSIGNAL, BROWNSCUSUMTRACKINGSIGNAL, TRIGGSSMOOTHEDTRACKINGSIGNAL, AUTOCORRELATIONTRACKINGSIGNAL | NOTRACKING |
TrackingSignalDataClass. type_: To set the
type of tracking signal calculation used for all algorithms, except the
Crostons method.
|
| Tracking Control Limit | BEST, H-W, MA, EWMA, LS | 0 < x <= 100 | 100 |
TrackingSignalDataClass. controlLimit_: Used
to set the control limit to trigger a tracking point. A value of 100
effectively stops the tracking signal detection.
|
Note: The system sets the tracking signal type, irrespective of the
specified setting based on these conditions:
- If the algorithm type is set to Moving Average or Least Squares, the tracking type is set to Autocorrelation. In this scenario, the control limit is also reset to 0.99.
- If the algorithm type is set to Holt-Winters or Crostons and the tracking signal type is set to Autocorrelation, the return value is set to Simple.
- For the previous period tracking system, the application resets the tracking signals.
When you call an engine to start the optimization run, that is,
bOptimise = true or
optimiseSmoothingCoefficientsLAHC(, use these parameters:
| Display Name | Used by (Algorithms) | Accepted range | Description |
|---|---|---|---|
| Optimization Idle | BEST, H-W | positive integer |
MainDataClass.OptimisationIdle_: The
Late Acceptance Hill Climbing (LAHC)
optimization heuristic parameter is used to
determine the number of iterations that can be performed, without requiring a
alternate solution, before the optimization is completed. This parameter is
also used determine the size of the fitness array and does not affect the
performance or the forecast result. Therefore, the use of the default value is
recommended.
|
| Optimization Idle Percentage | BEST, H-W | 0 < x <= 1 |
MainDataClass.
optimisationIdlePercentage_: The
Late Acceptance Hill Climbing (LAHC)
optimization heuristic parameter is used to
determine the number of candidate solutions that must be calculated
consecutively, when the solution is not saved to complete the optimization run.
This is the percentage of the total number of runs that can occur, based on the
incremental optimization and the model form. If this value is equal to one, the
optimization is used to determine the total number of runs that can occur based
on the incremental optimization and the model form. If this value is zero, a
single run is performed using the optimization starting values for the relevant
smoothing coefficients. The use of the default value is the recommended. As
this value reaches 1, the length of the run increases and therefore the
performance decreases. However, the possibility of reaching the required global
minimum value (the Best smoothing parameters) increases.
|
| Optimization Increment | BEST, H-W | positive decimal |
MainDataClass.
optimisationIncrement_: The
Late Acceptance Hill Climbing (LAHC)
optimization heuristic parameter is used to
determine the quantum of increase for the relevant Holt-Winters smoothing
coefficient, on an iteration of the optimization algorithm. Decrease this value
towards zero to increase the number of iterations, which in turn decreases the
overall optimization performance.
|
| Optimization Starting Value Alpha | BEST, H-W | positive decimal |
MainDataClass.
OptimisationStartingValueAlpha_: The
Late Acceptance Hill Climbing (LAHC)
optimization heuristic parameter is used to
determine the starting value of the level smoothing coefficient for the
Holt-Winters algorithm.
|
| Optimization Starting Value Beta | BEST, H-W | positive decimal | (MainDataClass. OptimisationStartingValueBeta_): The Late Acceptance Hill Climbing (LAHC) optimization heuristic parameter is used to determine the starting value of the trend smoothing coefficient for the Holt-Winters algorithm. |
| Optimization Starting Value Gamma | BEST, H-W | positive decimal |
MainDataClass.
OptimisationStartingValueGamma_: The
Late Acceptance Hill Climbing (LAHC)
optimization heuristic parameter is used to
determine the starting value of the seasonal smoothing coefficient for the
Holt-Winters algorithm.
|
When a debug output report from the engine is required, use these parameters:
| Display Name | Used by (Algorithms) | Accepted range | Description |
|---|---|---|---|
| Optimization Output | All | true or false |
MainDataClass. OptimisationOutput_: If
bOptimise = True, the parameter is used to start (true) or end (false) the
process to create an output file that contains entire result of the
optimization process. These results are written to the
SmoothingCoefficientOptimisationLAHCResults.txt file of the Tomcat folder at
the application's installation location. For example,
C:\Program Files\Infor\Sales and Operations Planning\
\Tomcat\.
|
| Output | All | true or false |
MainDataClass. Output_: The parameter is used
to start (true) or end (false) the creation of an output file detailing the
results from the Best algorithm or the specified algorithm. The output files
are PickingAlgorithmResults.txt and GivenAlgorithmResults.txt respectively. The
output files are written to the Tomcat folder at the application's installation
location.
|
| Tracking Output | All | true or false |
MainDataClass. TrackingOutput_: If Performance
Tracking = True, the parameter is used to start (true) or end (false) the
process to create an output file detailing the tracking signal and step change
process for all algorithms except Crostons. The results are written to the
TrackingResults.txt file to the Tomcat folder at the application's installation
location.
|
Note:
- If optional parameters are not defined for the engine's settings, the default calculation settings as specified in the optimization parameters table are used.
- Using the output option impacts the performance. It is recommended that you use this option only for testing or specific debugging cases.
Use these parameters to define an initial model fit:
| Display Name | Used by (Algorithms) | Accepted range | Description | Engine Default |
|---|---|---|---|---|
| Trim factor | Best and Classic Combining set to On | Positive Integer | Indicates the number of extreme
values to be removed when running the Trimmed Mean technique of Forecast
Combining. This represents a one-sided value that is when you remove the
first and last values, x =1. Note: This is applicable when Forecast Combining option is set to
On.
|
1 |
| Winsorizing factor | Best and Classic Combining set to On | Positive Integer | Indicates the number of extreme
values to replace when running the Winsizoring technique. This represents
a one-sided value that is when you remove the first and last values, x
=1. Note: This is applicable when Forecast Combining option is set to On.
|
1 |
| weighted AIC Maximum | Best and Classic Combining set to On | true or false | Indicates the maximum difference from the lowest AIC value for a technique to be included in the weighing calculation. This is used in Weighted AIC technique and applicable when Forecast Combining option is set to On. | 4 |
| Learning rate | Best and any Machine Learning option (ML Train-Test, ML Train-Test randomization, ML Cross-Validation, ML full dataset, ML full dataset randomization) is set to On. | Positive Decimal | Determines the errors that are built back into the
model during the calculation of the model coefficients. Note: This is
applicable when Forecast
Combining option is set to On.
|
0.001 |
| Epoch |
Best and any Machine Learning option (ML Train-Test, ML Train-Test randomization, ML Cross-Validation, ML full dataset, ML full dataset randomization) is set to On. |
Positive Integer | The number of loops of the entire data set that the
model performs to calculate the model coefficients. Higher the value of epoch, more accurate the model, but impacts the performance |
5 |
| Folds | Best and Machine Learning Cross-Validation option is set to On. | Positive Integer | The number of folds used in setting up
cross-validation data sets. Note: This is applicable when Forecast Combining option is set to
On.
|
5 |
| Random number seed | Best and (ML Train-Test randomization or ML full dataset randomization) set to On. | Integer | Possible values:
Note: This is applicable when Forecast Combining option is set to On.
|
-1 |