The explained variance of a principal component is its eigenvalue divided by the sum of all eigenvalues of the covariance matrix. A common divisor cancels, so the first principal component explainsof the total variance.
Articles by others on the same topic
There are currently no matching articles.