Prediction data unavailable for the reference. Change reference or perform a new evaluation.
Prediction data unavailable for current model evaluation. Please perform a new evaluation.
| Population stability index (PSI) |
{{uiState.driftState.driftResult.predictionDriftResult.populationStabilityIndex | mlMetricFormat: 3 | ifEmpty: '-' }}
|
|---|---|
| KS test |
{{uiState.driftState.driftResult.predictionDriftResult.ksTestPvalue | mlMetricFormat: 3 | ifEmpty: '-' }}
|
| Chi-square test |
{{uiState.driftState.driftResult.predictionDriftResult.chiSquareTestPvalue | mlMetricFormat: 3 | ifEmpty: '-' }}
|
| Current (threshold: ) | Current | Reference (threshold: ) | Reference | |
|---|---|---|---|---|
| Current | Reference | |||
Predicted {{currentClass}} (%) |
{{uiState.driftState.curPredValueCount[currentClass].pct}}Not available | {{uiState.driftState.refPredValueCount[currentClass].pct}}Not available |
Performance data are not available
Fugacity expresses the difference between the expected "ideal" data your model was trained on and the observed "real" data you are analyzing. We compare the proportion of samples predicted in each class when scoring on both the test and your input datasets.
This charts represents the probability density estimation for a given prediction class when scoring both the test dataset and your input dataset.
Visually different probability density estimations indicate high data drift.
Data unavailable for Reference.
Data unavailable for current model evaluation.
This chart represents the density estimation of the prediction values when scoring both the test dataset and your input dataset.
Visually different density estimations indicate high data drift.