Analytics (with PAI 5.0.0.1) > Available PAI Dashboards > Forecast Intelligence > Forecast Accuracy Intelligence > Forecast Accuracy Intelligence (Technical Specification)
Forecast Accuracy Intelligence (Technical Specification)
These are the mini-dashboards that comprise the Forecast Accuracy Intelligence dashboard:
• Forecast Data Overview — The Total SKU streams that were analyzed and their spread of tracking signal
• Forecast Feature Importance — All of the key features that are governing the accuracy of the forecast
• Profiles — The set of SKUs that have specific criteria that was predicted to be highly, and consistently, over or under forecasted, with very high confidence.
Forecast Data Overview
The data used to generate this mini-dashboard is based on the following:
Analytical Object
AO_PAI_ML_FAI_FcstDataOverview
Query Object
QO_PAI_ML_FAI_FcstDataOverview
Data Source
PAI Machine Learning
Cab File
ForecastAccuracyIntelligence.cab
Profiles
Used for building the Forecast Accuracy Intelligence dashboard
• query
Predecessor Jobs
None
Build Type
None
Build Frequency
None
• TS Range
Measure
Mathematical Function
SKU-Streams
count of (SKUStream) from sku stream table (PARTID+"_"+LOCID+"_"+FORECASTSTREAMID)
# (-1.0 : -0.5)
count of SKUs with tracking signal in range (-1.0 : -0.5)
# (-0.5 : 0.5)
count of SKUs with tracking signal in range (-0.5 : 0.5)
# (0.5 : 1.0)
count of SKUs with tracking signal in range (0.5 : 1.0)
% (-1.0 : -0.5) Expression
percentage of SKUs with tracking signal in range (-1.0 : -0.5)
% (-0.5 : 0.5) Expression
percentage of SKUs with tracking signal in range (-0.5 : 0.5)
% (0.5 : 1.0) Expression
percentage of SKUs with tracking signal in range (0.5 : 1.0)
Forecast Feature Importance
The data used to generate this mini-dashboard is based on the following:
Analytical Object
AO_PAI_ML_FAI_FcstFeatureImportance
Query Object
QO_PAI_ML_FAI_FcstFeatureImportance
Data Source
PAI Machine Learning
Cab File
ForecastAccuracyIntelligence.cab
Profiles
Used for building the Forecast Accuracy Intelligence dashboard
• query
Parameters
None
Predecessor Jobs
None
Build Type
None
Build Frequency
None
• Features
Measure
Mathematical Function
Importance
Average of Importance
Profiles
The data used to generate this mini-dashboard is based on the following two Analytical Objects:
Analytical Object
AO_PAI_ML_SFA_SKUErrorDetails
Query Object
QO_PAI_ML_SFA_SKUErrorDetails
Data Source
PAI Machine Learning
Cab File
MachineLearningForecasting.cab
Parameters
None
Predecessor Jobs
None
Build Type
None
Build Frequency
None
• ForecastStream
• HighRisk
• Location
• MethodName
• Part
Measure
Mathematical Function
Bias
Sum of Bias
Bias Value
Sum of (Bias * Price)
Average of MAPE
Average of RMSE
Average of MAD
Average of Forecast Accuracy
Average of Tracking Signal
Analytical Object
AO_PAI_ML_FAI_Profiles
Query Object
QO_PAI_ML_FAI_Profiles
Data Source
PAI Machine Learning
Cab File
ForecastAccuracyIntelligence.cab
Profiles
Used for building the Forecast Accuracy Intelligence dashboard
• query
Parameters
None
Predecessor Jobs
None
Build Type
None
Build Frequency
None
• Criteria
• Forecast Category
• Forecast Stream
• Location
• Part
• Profile Name
Measure
Mathematical Function
Members
Average of Members
Predicted Tracking Signal
Average of Predicted Tracking Signal
Predicted MSE
Average of Predicted MSE
Average of Tracking Signal
Average of MAPE
Average of MAD
Average of RMSE
Was this helpful?