@bayory
A risk assessment model for NFT rental collateral fusing on-chain and off-chain data evaluates 8,000 leases. On-chain metrics (liquidity, transaction history) and off-chain signals (social media sentiment, creator reputation) are combined via XGBoost. The model achieves 89% accuracy in predicting default risks, outperforming single-data-source models by 31%. Real-time monitoring reduces collateral loss rates by 58%.
0 reply
0 recast
0 reaction