| 基于神经网络的股票预测模型 |
| |
| 引用本文： | 乔若羽.基于神经网络的股票预测模型[J].运筹与管理,2019,28(10):132-140. |
| |
| 作者姓名： | 乔若羽 |
| |
| 作者单位： | 中国科学技术大学 统计与金融系,安徽 合肥 230026 |
| |
| 摘 要： | 针对股票市场的特征提取困难、预测精度较低等问题,本文基于深度学习算法,构建了一系列用于股票市场预测的神经网络模型,包括基于多层感知机(MLP)、卷积神经网络(CNN)、递归神经网络(RNN)、长短期记忆网络(LSTM)和门控神经单元(GRU)的模型。 针对RNN、LSTM和GRU无法充分利用所参考的时间维度的信息,引入注意力机制(Attention Mechanism) 给各时间维度的信息赋予不同权重,区分不同信息对预测的重要程度,从而提升递归网络模型的性能。上述模型均基于股票数据进行了优化,基于上证指数对各类模型进行了充分的对比实验,探索了模型中重要变量对性能的影响,旨在为基于神经网络的股票预测模型给出具体的优化方向。
|
| 关 键 词： | 股票预测 深度学习 神经网络 注意力机制 |
| 收稿时间： | 2018-08-10 |
Stock Prediction Model Based on Neural Network |
| |
| QIAO Ruo-yu.Stock Prediction Model Based on Neural Network[J].Operations Research and Management Science,2019,28(10):132-140. |
| |
| Authors: | QIAO Ruo-yu |
| |
| Affiliation: | Department of Statistics and Finance, University of Science and Technology of China, Hefei 230026, China |
| |
| Abstract: | To address the issues corresponding to the difficult feature extraction and low prediction accuracy in the stock market, this paper proposes a series of neural network-based models for stock market forecasting based on deep learning, which refers to multi-layer perceptron, convolutional neural network, recurrent neural network, long-and-short-term memory network and gated recurrent unit, respectively. In consideration of the case that RNN、LSTM and GRU models can not make full use of the referred information in time dimension, this paper introduces the attention mechanism to give different weights to the information of each time dimension which can distinguish the importance of different information to the prediction, and then improve the performance of the recursive network model. All the models are optimized with respect to the stock data. Extensive experiments on the Shanghai Stock Index demonstrate the effectiveness of the proposed models. In addition, the comparisons between different models further explore the effect of the important parameters in the models. The purpose of this paper is to give the optimization direction of the stock forecasting models based on neural network. |
| |
| Keywords: | stock prediction deep learning neural network attention mechanism |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《运筹与管理》浏览原始摘要信息 |
|
点击此处可从《运筹与管理》下载全文 |
|
