Application of Improved Principal Component Analysis Algorithm in Course Construction
|School||Harbin Engineering University|
|Keywords||multivariate statistical analysis the principal component analysis course construction evaluation index|
Course construction is a combination that is research and teaching of the institutions of higher education, and is an important indicator of school level. It is an important means to implement course assessment, which can strengthen graduate education, promote scientific research, and improve the quality of personnel training. To solve the question of how to pick up the main index factor in the course construction, in this paper, the principal component analysis methods in the multivariate statistical analysis methods is introduced, and quantitative analysis of preliminary is given for assessment of course construction.Principal component analysis is a technique used for reducing the dimension in the multivariate statistical analysis, and it’s the most widely used method for comprehensive assessment. However, basing of no loosing of original information, the traditional principal component analysis is sensitive to the original data, and the effect of reducing the dimension is not obvious, and it increases the complexity of the question, finally it’s led to a lower efficiency of principal component analysis. In this paper, the treatment of the original target and the Determination of indicators weights are taken as the direction of improving the Principal component analysis, by the Simulation experiment, We proposed an improved principal component analysis method. It’s proved the effect of reducing the dimension is obvious used the improved principal component analysis.In this paper, we improve the principal component analysis and apply to the course construction. According to the ideas and principles of index system for the assessment of course construction, the evaluation index system which apply to the data of this paper is established. Basing on the data weighted equalization treatment before assessing the data, we asses the data index factor with the improved principal component analysis method. The experimental results showed that the improved F principal component analysis is an effective method of assessment and analysis. This method has promising prospect.