A PSO-Based Ensemble Model for Peer-to-Peer Credit Scoring

We propose a multi-classifier ensemble model based on particle swarm optimization (PSO) for the evaluation of personal credit risk in peer-to-peer (P2P) lending platforms. In the proposed method, we consider the differences and complementarity of the base classifiers' performance and use PSO to optimize their weights. Experimental results show that our proposed P2P personal credit scoring model outperforms both single and other benchmark ensemble models. Among the examined model variants, the ensemble model based on PSO with 100 particles is the best.