Liangzi AUTO: A Parallel Automatic Investing System Based on GPUs for P2P Lending Platform

Abstract

Nowadays, automatic investing becomes more and more popular in P2P platforms. For the sake of satisfying the lenders' better experiences and offering more flexible automatic investing strategy, in this paper we propose Liangzi AUTO-a parallel automatic investing system which can dynamically decide the investing sequences according to four factors at the moment of starting recruiting of a loan project. We design a dynamic scheduling algorithm for automatic investing and parallelize the scheduling algorithm to exploit Liangzi AUTO with assistance of modern GPUs. Experimental results show that we achieve a maximum speedup of 37.40X.