@msonali30022814
Default confidence is the threshold for that class (Output value) and the predicted confidence is based on the information in the Input fields and how confident it on the prediction based on this input. If the prediction confidence is higher than the default confidence, it is above the threshold and it will populate the predicted value on the record. If the opposite and below the threshold, it will be "Skipped", and the predicted value is not populated on the record. The discrepancy is caused by the values in the Input fields for the prediction, and so the predicted confidence will always vary. However, the default confidence is a static value, and only changed when the solution is retrained with newer data and so new thresholds (default confidence) are generated as Precision Coverage Lookup records, and typically close to the previous default confidence value, depending how much the data changed in the solution training dataset.
