The weighted least squares objective isIts gradient isSince is positive diagonal and has full column rank, is positive definite. The unique stationary point is therefore the unique minimizer:
Articles by others on the same topic
There are currently no matching articles.