1. A transformation applied uniformly to each element in a set of data so that the set has some specific statistical property. For example, monthly measurements of the rainfall in London might be normalised by dividing each one by the total for the year to give a profile of rainfall throughout the year. 2. Representation of a floating-point number so that its mantissa's left-most digit is non-zero. If the leftmost fraction digit are zeros, the number is said to be unnormalised. Unnormalised numbers are normalised by shifting the fraction left, one digit at a time, until the leftmost digit is nonzero and reducing the exponent by the number of shifts. 3. database normalisation. |