DAVIDON-FLETCHER-POWELL

Davidon-Fletcher-Powell (DFP) update - This is a way to update an approximation of an inverse hessian , used for unconstrained optimization. Using the notation in the DFP method , the update is:

 pp' Hqq'H H' = H + --- - -----. p'q q'Hq 

Note: pp' is a rank 1 matrix (the outer product of the vector p with itself).