A radix numeration system in which the location of the decimal point is indicated by an exponent of the radix; in the floating-point representation system, 0.0012 is represented as 0.12-2 where -2 is the exponent. Synonym: floating-point representation system.