基于深度学习的股票市场价格预测

论文总字数：18290字

目 录

一 概论 1

1.1本课题的背景及意义 1

1.2本课题的国内外研究现状 1

1.2.1 国外研究结果 1

1.2.2 国内研究结果 2

1.3本课题研究的主要内容 2

二 股票预测的概念和理论 2

2.1 股票市场的基本概念 2

2.2 股票价格预测方法 3

2.2.1 基本面分析法 3

2.2.2 技术面分析法 4

三 神经网络理论及模型 4

3.1 人工神经网络 5

3.1.1 人工神经网络介绍 5

3.1.2 人工神经网络模型构建 5

3.2 BP,RNN神经网络理论研究 10

3.2.1 BP神经网络理论和模型 10

3.2.2 RNN神经网络理论和模型 13

3.3 LSTM神经网络理论研究 15

3.4 神经网络框架Tensorflow的介绍 20

3.4.1 Tensorflow框架的背景 20

3.4.2 Tensorflow的基本概念 21

四 实验预测及结果分析 23

4.1样本数据 23

4.2模型设计 23

4.3实验过程 24

4.4实验结论 28

五 结语 28

参考文献 29

致谢 30

基于深度学习的股票市场价格预测

梁栋

,China

Abstract:The stock market is an important part of a country's economy. However, the instability of the stock market itself can even bring about an irreversible financial turmoil. In order to minimize this risk, this paper uses a neural network model for deep learning.The neural network model is known for its accurate and reliable data prediction advantages, as compared to the traditional statistical prediction methods.

Based on the analysis of the importance of the stock market and the analysis of prediction methods, this paper uses the Lstm neural network as a predictive model. The advantage of the Lstm neural network is that it can effectively predict the time series. Based on the famous open source framework Tensorflow, a neural network model was established. The opening price, the closing price, the lowest price of the day, the highest price of the day, the trading volume, the total price of the transaction, the rate of change, and the highest price of the next day are the input parameters, and the results are predicted continuously by increasing the number of trainings and changing the learning rate.

According to the experimental results, the prediction of the weekly stock price deviation rate can be controlled very low, which shows that the use of neural networks for stock market price forecasting is feasible. When the neural network model algorithm matures and the model is more perfect, the prediction The effect will become very impressive.

Keywords: Artificial intelligence Stock prediction Deep learning Neural network

一 概论

1.1本课题的背景及意义

股票市场是当今发达国家以及发展中国家的经济大梁，虽然它不是国民经济的命脉，但就发展态势而言，已经走入人们的日常生活中。纵观当下，阿里，腾讯等互联网大公司早已投入到股票市场中，阿里在美国的上市使得阿里公司本身的市值提高了不少，这说明股票市场对于一个企业的发展更多的是一次机遇。

我国的股票市场在改革开放以来以雨后春笋地形式急速增长，中国目前的上市公司已经达到了2000家，其中不乏像万达这样的财团，这也预示着股票市场的蓬勃发展，再加上政府的大力支持，一定会成为国民经济的顶梁柱。

剩余内容已隐藏，请支付后下载全文，论文总字数：18290字