A Formally-Proved Algorithm to Compute the Correct Average of Decimal Floating-Point Numbers | IEEE Conference Publication | IEEE Xplore