Multi-dimension Data Sources
TrustScore constructs a credit scoring model based on the user base information, multi-platform information of credit application, user behavior preference, asset level, e-commerce consumption, credit history and other information inside and outside the TrustDecision. Rich data type is the effective guarantee of scoring effect.
Sustainable Risk Protection
TrustScore has good feedback effect in many customer tests and practical use, with high differentiation of score results, model KS>=0.35, high scoring stability, and modelPSI<=0.01.
Advanced Machine Learning Algorithms
TrustScore provides customers with accurate and reliable personal credit evaluation solutions based on cutting-edge machine learning algorithms, including XGBoost, LightGBM and other algorithms.