BFGS Method

A quasi-Newton optimization algorithm that iteratively builds and updates an approximation to the inverse Hessian using gradient differences, enabling efficient unconstrained minimization with superlinear convergence near a local minimizer without computing the exact Hessian.