Font Size: a A A

Multi-dimensional Statistical Data Quality Evaluation Method Based On Improved TOPSIS Research

Posted on:2022-08-16Degree:MasterType:Thesis
Country:ChinaCandidate:S YuFull Text:PDF
GTID:2480306575482284Subject:Mathematics
Abstract/Summary:PDF Full Text Request
Data quality is very important for statistical data analysis.How to choose a reliable research method,an effective evaluation of data quality,for all levels of government and society to provide detailed and reliable statistical information is not only beneficial to the right judgment of information demanders,and help them to make effective decisions,but also the era of big data quality evaluation and an essential link.Therefore,it is necessary to conduct an in-depth study on statistical data quality evaluation methods,and the main work is as follows:(1)In the accuracy dimension evaluation,considering the complexity of the relationship between economic variables,if we simply use non-parametric regression model or parametric regression model to deal with it,part of the information will be lost,resulting in a large error between the estimated result and the actual value.Therefore,the use of semi-parametric model in this paper can avoid the shortcomings of general simple parametric non-parametric model and has the advantage of better fitting effect,and the selection of semi-parametric model for evaluation accuracy has a good effect.(2)Aiming at the comprehensive evaluation of the five evaluation dimensions of completeness,accuracy,timeliness,effectiveness and consistency,the TOPSIS method is adopted,which does not have too many restrictions on the sample size and index of evaluation size and is very applicable.Moreover,the cloud-based method has fuzziness and has more advantages in distinguishing data of this and that.The cloud model is innovatively adopted to improve the weight of TOPSIS method,and the difference between data is considered on the basis of the expected weight of the original cloud model,so as to realize the weight improvement.Expectation is combined with entropy to determine the weight method is proposed,the empirical analysis shows that the traditional TOPSIS,entropy weight method,entropy and entropy of empowerment and cloud model of cloud model,the combination of empowerment TOPSIS to evaluate three methods of data quality,improve the empowerment of the method is more advantageous to realize the sorting and evaluation on the data quality,applicability and strengthened.Figure 12;Table 25;Reference 52...
Keywords/Search Tags:quality of statistical data, TOPSIS comprehensive evaluation method, semi-parametric model, cloud model
PDF Full Text Request
Related items