Comparison and Empirical Analysis between Principal Component Analysis and Factor Analysis
Lu Ning Xu, Yingying Zhang
Abstract
Open-access reader
Lu Ning Xu, Yingying Zhang
Abstract
Open-access reader
We carry out an empirical analysis based on the study of the principal component analysis and the factor analysis, the data is in the 2012 China Statistical Yearbook [ 1 ] with 31 provinces and 9 major economic indicators.First by comparing three kinds of methods of factor analysis, we find out that the principal factor analysis method has the minimum sum of squared errors.Then we use the principal component analysis and the principal factor analysis to extract two principal components and two principal factors from the economic indicators.Third, we calculate principal component scores, factor scores, and their comprehensive scores of the provinces under two methods.Finally, we find that the rankings of the provinces under the two methods are different, but the overall trends are consistent, and the trends are also consistent with the actual situations.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We carry out an empirical analysis based on the study of the principal component analysis and the factor analysis, the data is in the 2012 China Statistical Yearbook [ 1 ] with 31 provinces and 9 major economic indicators.First by comparing three kinds of methods of factor analysis, we find out that the principal factor analysis method has the minimum sum of squared errors.Then we use the principal component analysis and the principal factor analysis to extract two principal components and two principal factors from the economic indicators.Third, we calculate principal component scores, factor scores, and their comprehensive scores of the provinces under two methods.Finally, we find that the rankings of the provinces under the two methods are different, but the overall trends are consistent, and the trends are also consistent with the actual situations.
Key concepts: Principal component analysis, Factor (programming language), Statistical analysis, Statistics, Correspondence analysis, Factor analysis, Principal (computer security), Computer science