Improving the effectiveness of floating point arithmetic | IEEE Conference Publication | IEEE Xplore