ABSTRACT
In [5], the authors developed an iterative scheme for approximating solutions of certain classes of nonlinear integral equations. In the numerical experi
ments, wavelet bases worked very well with the method. However, problems did occur when trying to approximate the solution at the endpoints of the in terval where the solution was desired. In part, this error can be attributed to Gibbs’ phenomenon in our summation techniques. Moreover, in the classes where both the kernel and the solution are nonnegative it would seem natural to use positive basis functions.
