I calculate a scores of different sets of data like this:
$$scores_j = avg(x_{j,i} * a_i)$$
where $a_i$ are constant and $x_{j,i}$ change between data sets. Let say I have as a result:
$$scores_1 = (10,20,30)$$ $$scores_2 = (10,20,40)$$
then I rescale them to the range of 0 .. 100
$$scores_1 = (0,50,100)$$ $$scores_2 = (0,33,100)$$
My question is by comparing 33 =?= 55, m'I comparing Apples to Apples OR Apples to Oranges ? /the same for 0 and 100/
If they are not comparable, how can I make them comparable ?
lets say X is a preference from 0 .. 1 of feature F and A is support of feature F for item M , 0 .. 1