EMD2FNN: A strategy combining empirical mode decomposition and factorization machine based neural network for stock market trend prediction

作者：

Highlights：

• We propose an improved neural network model to predict the stock prices.

• The empirical mode decomposition and factorization machine are used in our approach.

• The empirical mode decomposition helps overcome the non-stationarity of stock price.

• Factorization Machine helps grasp the nonlinear interactions among the inputs.

• The real data sets are used to demonstrate the accuracy of the new approach.

摘要

•We propose an improved neural network model to predict the stock prices.•The empirical mode decomposition and factorization machine are used in our approach.•The empirical mode decomposition helps overcome the non-stationarity of stock price.•Factorization Machine helps grasp the nonlinear interactions among the inputs.•The real data sets are used to demonstrate the accuracy of the new approach.

论文关键词：Empirical mode decomposition,Factorization machine,Neural network,Stock market prediction,Profitability

论文评审过程：Received 1 April 2018, Revised 2 July 2018, Accepted 29 July 2018, Available online 30 July 2018, Version of Record 14 August 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.07.065