linear algebra notation and regression models
I am new here; hope that I am posting in the right forum.
What does the expression
mean? .... where b is a vector, P is a matrix and l is a skalar. I don't know what b' means.
It's the penalty term from penalized maximum likelihood estimation in regression analysis. b is the vector of regression coefficients, l is an overal penalty factor, and the matrix P is used to scale penalties on individual coefficients bi, e.g. by their standard deviations. However, I don't understand why you'd need a matrix for the latter; why not simply define a vector of individual scales. In fact all examples that I saw so far used a diagonal matrix.
Thanks a lot for help !!
monalisa (biologist, learning advanced statistics)