摘要
针对数据分析精确性的评估方法和发展趋势,说明大数据精确性还存在一定缺陷。利用IBM SPSS Statistics 20.0软件,对安徽省近十年（2004~2013年）的国民生产总值建立非线性回归与简单线性回归的模型进行比较,用拟合结果定量评估大数据分析的精确性,用拟合优度数值说明分析方法产生的差异性;同时利用非线性与线性模型同时预测了2014年的GDP数值,并与2014年实际GDP数值进行比较,再次说明了精确性分析的差异与意义,由此提出大数据分析精确性的难点和建议。
Based on the data analysis accuracy evaluation methods and the development trend, defects still exist in researching data accuracy. Using the IBM SPSS Statistics 20.0 software to establish two models that are nonlinear regression model and simple linear regression model to compare which is more accuracy with the data of Anhui province GDP（2004-2013）.Then, fitting the results to evaluate the accuracy and the difference of the data; at the same time, using nonlinear and linear model to predict the 2014 GDP value and compare it to the real GDP in 2014, the result shows the difference and the meaning of accuracy analysis of the data. Finally, the paper puts forward some suggestions on the difficulties in the accuracy analysis of big data.
出处
《信息技术与标准化》
2016年第7期29-32,37,共5页
Information Technology & Standardization
基金
北京市科技计划项目"数字科技档案自动化与利用服务系统设计研发"
项目编号:Z151100003215042
关键词
大数据
数据精确性评估
非线性回归
线性回归
big data
evaluation of data accuracy
nonlinear regression
linear regression
