Floating-Point Computation of Functions with Maximum Accuracy | IEEE Journals & Magazine | IEEE Xplore