A P2P Online Lending Agency Risk Identification Approach Based on Reinforcement Learning

This paper proposes a risk identification method for P2P online lending institutions based on improved reinforcement learning model, which regards risk identification task as a kind of text classification. In order to solve the problem of small-scale data and data imbalance of P2P online lending institution for text classification using existed supervised learning methods, this paper adopts Reinforcement Learning model. Unfortunately, existed Reinforcement Learning model are usually not consider the global information. Hence this paper continues to put forward three improved Reinforcement Learning models to consider global information by attention weight and information value. Experimental results have shown that our proposed models have better performance than the basic model.