- Partition Strategy of Smart Predict in SAP Analytics Cloud (SAC)
- SAP Analytics Cloud
- Smart Predict
Reproducing the Issue
[Enter your Reproducing the Issue here...]
[Enter your Cause here...]
Partition Strategy is also called "Cutting Strategy" in Predictive Analytics. Refer to: Partition Strategy
A partition strategy is a technique that decomposes a training dataset into distinct subsets.
Differenct from Predictive Analytics, in Smart Predict, training dataset is decomposed into TWO distinct datasets:
- A training subset (Estimation dataset) - Generate different models. The models generated at this stage are hypothetical.
- A validation subset - Select the best model among those generated using the training subset, which represents the best compromise between perfect quality and perfect robustness.
The partition is performed as follows:
- The row selection is random.
- The training subset contains 75% of the input rows.
- The validation contains 25% of the input rows.
How the partition strategy works when training a Predictive Model
The following graphics summarize what's happen when you click train:
Training dataset is selected from input dataset - randomly 75% of input dataset. Differenct trial version of model generated. Then validation dataset - 25% of input dataset is used to validate the trial versions, and select the best Predictive model. Refer to: Training a Predictive Model
PA, Infinite insight, test sub-set, training a model, classification, time series, regression , KBA , LOD-ANA-PR , SAP Analytics Cloud – Predictive (BOC) , BI-RA-PA , Predictive Analytics , How To