As a result, all the data in the training data set, 70 percent of the data in the mining structure, is used for cross-validation.
This process is repeated until Analysis Services has created and tested the specified number of models.
The data that you specified as being available for cross-validation is distributed evenly among all partitions.
, meaning the probability that the predicted state is correct.
If the predict probability exceeds the accuracy bar, the prediction is counted as correct; if not, the prediction is counted as incorrect.
Note You can set a value of 0.0 for the threshold, but the value is meaningless, because every prediction will be counted as correct, even those with zero probability.