摘要随着大数据时代的到来，数据资产管理也日益受到人们尤其是企业的关注。本文分析了数据资产管理过程中十分关键的数据治理部分的大体框架，并选取股票市场数据作为管理对象，重点针对股票的投资分析预测，设计并实现股票数据资产管理系统，对现有的股票信息的系统管理做了补充，主要内容如下：对数据治理的流程与数据流概念做了介绍，接着对股票市场的常用术语与基本面分析做了解释，然后通过TuShare获取并存储源数据，为了便于数据利用进行了数据清洗，具体是针对完整数据记录的离散化与归一化，最后利用神经网络对标准化的数据进行预测分析，通过分析对比讨论结果，并完成相关部分的可视化。42785

关键词  数据治理  数据流  数据清洗 神经网络

毕业论文设计说明书外文摘要

Title  Design and Implementation of Data Asset Management System                                            

Abstract

With the advent of the age of big data, data asset management has attracted increasing attention from people, especially enterprises. This paper analyzes the general framework of data governance which is the key process of the data asset management, and selects the stock market data as the object of management. Meanwhile, this paper mainly focuses on analysis and prediction of stock investment, and design and implementation of the stock data asset management system. Finally, some supplement is made to existing stock information system. The main contents are as follows. First, we provide a short introduction to the process of data governance and the concept of data flow. Then we make some explanation upon the common terms of stock market and the fundamentals of analysis. Next, through TuShare, source data can be obtained and stored. To make data utilization more convenient, method of data cleaning is used. Specifically, we make it the full data records discrete and normalized. Finally, we use neural network to conduct prediction and standardized data analysis, through analysis to and comparison among obtained results we complete the visualization process for relevant parts of them.

Keywords  Data governance  Data flow  Data cleaning  Neural network

目   录

1  绪论 2

1.1  本文研究的背景与意义 2

1.2  国内外研究现状 3

1.3  本文研究简述与解决方案 3

1.4  本章小结 4

2  数据资产管理系统中的基本概念与原理 4

2.1  数据流 4

2.2  金融市场概述 5

2.3  神经网络概述 6

2.4  梯度下降 6

2.4.1  线性回归 7

2.4.2  错误函数选用平方和的概率解释 8

2.4.3  梯度下降 9

2.4.4  有关梯度下降的其他讨论 10

2.5  BP神经网络 11

3  源数据的获取与基本操作 14

3.1  TuShare 14

3.2  选取的源数据 14

3.3  数据的获取与存储