Factor analysis-Model parameter estimation method

Factor load proof A and characteristic variance the method of proving D has principal component method, the principal is the factor FA and the maximum likelihood method.

Factor load Matrix: The art of variable factor expression in each element, the degree of influence of the common factor of expression extraction on the original variable

function: The linear combination of the original index variables can be obtained by the factor load matrix;

Example: such as X1=A11*F1+A12*F2+A13*F3, where X1 is the indicator variable 1,a11, A12, A13 are factor loads with variable X1 on the same line respectively

Factor score Matrix: Indicates the relationship between the index variables and the extraction of common factors, and the high score in a common factor indicates that the index is closely related to the public factor.

Effect: The linear combination of common factors can be obtained by Factor score proof

Examples: F1, F2, F3 are the common factors for extraction, and the linear combination of common factors, such as f1=a11*x1+a21*x2+a31*x3, are obtained by Factor score matrix.

Principal Component method:

Feature vectors and feature roots:

If A is a matrix, X is a non-zero vector that makes Ax=ax, where A is a quantity (which can be 0), then A is a eigenvalue (root) of a, and X is a eigenvector corresponding to a.

———— to Be Continued

Factor analysis-Model parameter estimation method