In the last chapter, we examined the numerical method known as Newton's method.
We established that one of the major disadvantages of this method was that that J(x) and
its inverse must be computed at each iteration. We, therefore want to avoid this problem.
There are methods known as Quasi-Newton methods, in which Burden and Faires in [3]
describe as methods that use an approximation matrix that is updated at each iteration
in place of the Jacobian matrix. This implies that the form of the iterative procedure
for Broyden's method is almost identical to that used in Newton's method. The only
exception being that an approximation matrix Ai is implemented instead of J(x). With
that said the following equation is derived:
x(i+1) = x(i)