On the number of principal components: A test of dimensionality based on measurements of similarity between matrices

摘要: An important problem in principal component analysis (PCA) is the estimation of correct number components to retain. PCA most often used reduce a set observed variables new lower dimensionality. The choice this dimensionality crucial step for interpretation results or subsequent analyses, because it could lead loss information (underestimation) introduction random noise (overestimation). New techniques are proposed evaluate PCA. They based on similarity measurements, singular value decomposition and permutation procedures. A simulation study conducted relative merits approaches. Results showed that one method RV coefficient very accurate seems be more efficient than other existing