CP
for each control point, we give 8 scores. 7 and overall
- Rules-based | Pass/Fail (0 or 100%)
- Expert Rules | Low/Medium/High (expert judgment)
- Statistical | 0~100% (statistical model)
- Machine Learning | 0~100% (data learning)
CP Weight Determination Criteria
- Audit importance
- False positive possibility
- Correlation with actual errors/fraud
CPs Recommended for Adjustment: CPs that require weight adjustment to be effective:
- Material Value: "Material amount" threshold varies by client
- Manual Entry: Manual entry patterns differ by company
- Suspicious Keywords: Suspicious keywords differ by industry
- POP-related: Need to adjust period-over-period variance thresholds
CPs where parameter settings are important when creating synthetic data
Not Recommended for Adjustment
Recommend maintaining default values for the rest
support.mindbridge.ai
https://support.mindbridge.ai/hc/en-us/sections/360010565994-Control-points-and-algorithms
Seonglae Cho