您的浏览器禁用了JavaScript(一种计算机语言,用以实现您与网页的交互),请解除该禁用,或者联系我们。 [TIPDM]:第三届挑战赛B6-基于数据挖掘技术的市财政收入分析预测模型 - 发现报告

第三届挑战赛B6-基于数据挖掘技术的市财政收入分析预测模型

2015-11-23 TIPDM 匡露
报告封面

全国大学生数据挖掘竞赛 优秀作品 作品名称:基于数据挖掘技术的市财政收入分析预测模型荣获奖项:二等奖作品单位:佛山科学技术学院作品成员:张永驰钟建锐王金茹指导教师:戎海武 广州财政收入分析与预测 摘要:本文对影响广州市财政收入各种因素进行了分析,并建立了各种经济指标预测的数学模型。 针对问题一,首先用相关性分析与多元回归分析的方法,分析了财政收入和增值税、营业税、企业所得税、个人所得税等四种税的相关性,并建立了相应的多元回归分析模型。SAS软件计算结果表明,四种税收都与财政收入有较强的相关性,影响财政收入的主要税种是增值税和企业所得税。其次,建立了增值税与商品进口总值、地区生产总值、工业增加值、批发零售业零售额、工业增加值占GDP比例、批发零售业等9个因素的多元回归分析模型,结果表明影响增值税的关键性因素是工业增加值;建立了营业税与公路货运量、公路客运量等9个因素的多元回归分析模型,影响营业税税的关键性因素是第三产业增加量;建立了企业所得税与第二产业增加值、第三产业增加值等10个因素的多元回归分析模型,影响企业所得税的关键性因素是城市商品零售价格指数;建立了个人所得税与城市居民年人均可支配收入、城镇单位职工年平均工资等7个因素的多元回归分析模型,影响个人所得税的关键性因素是地区生产总值和地方财政收入。最后,建立了财政总收入与地区生产总值和第一、二、三产业产值的回归模型,第三产业是影响地区生产总值的主要因素。 针对问题二,用时间序列分析的方法和曲线专家工具建立了财政总收入、地区生产总值、增值税等四种税收、第一、二、三产业产值、政府基金性收入以及从化市、增城市和市区的财政收入等指标的预测模型,结果表明增值税的增长较快、第三产业的增长速度快占地区生产总值的比例高于60%、广州市区的财政收入依然占主要部分。 针对问题三,给出了提高财政收入的建设性建议。 本文给出的回归模型及时间序列模型简单明了,精度高,容易用计算机软件实现。用相关分析及回归分析方法找到的影响广州财政收入的主要因素与广州市实际经济状况吻合,用时间序列分析方法给出的预测指明了财政收入及产业产值的发展趋势。 关键词:广州财政收入,因素分析,回归分析,时间序列分析,预测 Financial analysis and forecast of Guangzhou City Abstract:A brief description of the abstract This passage analyzes the factors that affect the financial revenue of Guangzhou, andestablishes the corresponding forecasting model. For question one, the methods of correlation analysis and multiple regression analysis areused to analyze the correlation between financial revenue and value added tax, business tax,corporate revenue tax and personal revenue tax, firstly. Meanwhile thecorresponding multipleregression analysis model is established.SAS software calculation results show that four kinds oftax have a strong correlation with the financial revenue, and the main taxes that affect the financialrevenue are the value added tax and enterprise income tax.Secondly, we establish a multipleregression analysis model of value added tax with commodity import value, regional GDP,industrial added value, wholesale and retail sales, industrial increase value accounted for theproportion of GDP, wholesale and retail trade.The results show that the key factor of the valueadded tax is the added value of industry.We establish multiple regression analysis model of thebusiness tax with road freight, highway passenger transport volume, the construction industryincrease value, the industrial added value, real estate development investment of the wholesociety,the whole society residential investment,construction industry total output value,accommodation and catering industry retail sales, business revenueof restaurant above thelimitation.The results show that the key factor of the value added tax is the added value ofindustry.We establish a multiple regression analysis model of the business tax with road freight,highway passenger transport volume, the construction industry increase value, the industrialaddedvalue,real estate development investment of the whole society,the whole societyresidential investment, construction industry total output value, accommodation and cateringindustry retail sales and business revenue of restaurant above the limitation.The results show thatthe key factor of the business tax tax is the third industry increase.We also establish the multipleregression analysis model of the personal revenue tax and urban residents, people can bedisposable revenue, staff and workers in urban units year of average wage, the balance of savingsdeposits of urban residents, GDP, the second industry added value, urban non private unitsemploying number and local financial revenue.The results show that the key factor that affects theindividual revenue tax is the regional GDP and the local financial revenue.Finally, we establish themultiple regression model of the total revenue, GDP and the first two or three industry output value.The results show that the third industry is the main factor to affect the nag of GDP.To solve question two, we use time series analysis method to established the prediction model of total financial revenue, GDP, value added tax and other four kinds of taxes, the first,second and third industry output value, government fund revenue , Conghua, Zengcheng City andurban financial revenue and other indicators.The results showed that the growth of value addedtax grew rapidly.Third the industry's growth rate is fast,becoming the main part of GDP.The speedof development of Guangzhou city is still high.To solve the problem three, some suggestions are given to improve the financial revenue. The regression model and the time series of the model are simple and easy to be realized by computer software.The main factors affecting the financial income of Guangzhou city are in theidentical with the actual economic situation of Guangzhou city by correlation analysis andregression analysis.The forecast points out the development trend of the financial revenue and theindustry output value by the time series analysis method. Key words:Guangzhou financial revenue,factor analysis,multiple regression analysis,timeseries analysis, forest 目录 1.引言.........................................................