The Davidon–Fletcher–Powell (DFP) formula is an algorithm used in optimization, specifically for finding a local minimum of a differentiable function. It is part of a family of quasi-Newton methods, which are used to approximate the Hessian matrix (the matrix of second derivatives) in order to perform optimization without having to compute this matrix explicitly. The DFP algorithm is particularly known for its ability to update an approximation of the inverse Hessian matrix iteratively.
Articles by others on the same topic
There are currently no matching articles.