Appearance
OperationRiskPredictionOutput
Type: Object
ML-predicted risk for an operation from the outcome classifier — flags operations likely to trigger audit warnings before problems manifest
Fields
| Name | Type | Args | Description | Deprecated |
|---|---|---|---|---|
| confidence | Float | XGBoost classification confidence (0–1). Null when not available. | ||
| operationId | String! | Operation this prediction applies to | ||
| predictedOutcome | String! | Predicted audit outcome class: pass, warning, or critical | ||
| topRiskFactors | [MLRiskContributionOutput!] | Top contributing features that drove the prediction (SHAP values). Positive values push toward warning/critical. Null when model does not support contributions. |