Enterprise AI Administration
Fraud Risk Classifier
Model Details · Telecom AI Model Registry
Model Details
Fraud Risk Classifier
Algorithm: XGBoostAPPROVED DESIGN
- Code
MDL-FRAUD-RISK- Version
- v2.1
- Owner
- Head of Fraud Management
- Governance owner
- Model Risk Committee
Governance configuration shown for review and validation. Read-only.
Business purpose
Scores subscribers and events for SIM-box, recharge and roaming fraud risk.
Production-recommended design specification. This record does not mean the model has been trained, fitted, deployed or validated.
Technical specification
- Model family
- Gradient Boosted Decision Trees
- Algorithm
- XGBoost
- Implementation library
- XGBClassifier
- Model type
- Binary / Multiclass Classification
- Explainability method
- SHAP TreeExplainer
Implementation governance
- Governance status
- APPROVED DESIGN
- Implementation status
- MODEL DEVELOPMENT REQUIRED
- Validation status
- VALIDATION REQUIRED
- Lifecycle stage
- Design
- Current performance
- Not yet measured
- Refresh cadence
- Daily
- Last reviewed
- 20 Aug 2026
- Next review due
- 20 Nov 2026
- Historical sample reference
- Precision at top 1% 0.81 — earlier sample figure, kept for provenance only; not a validated model result.
Business
- Mission area
- Fraud & Revenue Assurance
- Decision journey
- fraud-revenue-assurance
- Target variable
- fraud_confirmed_flag
- Output
- Fraud risk score (0–1) per subscriber-day
Input features
Governed variables, in feature-rank order
- 1.fraud_anomaly_scoreUnsupervised anomaly score on calling and recharge behaviour (0–1).
- 2.duplicate_recharge_countRecharges reusing a voucher or repeating within 60 seconds, last 7 days.
- 3.Roaming usage flagAny roaming usage in month
- 4.usage_pattern_deviationDeviation of current usage from the subscriber's 90-day profile (z-score).
Related Intelligence
Linked governed records
KPIs (2)
Business rules (2)
Formulas (0)
None recorded.
Source mappings (0)
None recorded.
Simulation models (1)