6.1.1 Classification of Unconstrained Minimization Methods
Several methods are available for
inte an unconst
ined minimization 9
d into two broad categories as direct setmization Methods
Descent Methods?
Steepest descent (Cauchy) method
Fletcher-Reeves method
Newton's method
Marquardt method
Quasi-Newton methods
Davidon-Fletcher-Powell metho!
Broyden-Fletcher-Goldfarb-Shannoy method
ie te
is and descent. methods as indicated in Table 6.1, The direct search
function values but not the partial deriva-
z ‘imum and hence are often called the
ids, The direct search methods are_also_known_as.zeroth-
(ee meshed since they use zeroth-order derivatives of the function. These
‘Chol ae.most suitable for simple, problems involving a.relatively_sroall
voor of-vatiables.—These methods are, in general, less efficient than the
scent methods. The descent techniques require, in addition to the function