Perceptual spaces defined by independent features

In the IFF, a distance for each feature is calculated independently and weighted, and then the total target cost is calculated as a function of these, fit the original Hunt and Black paper, this is given as 

Ui is a unit in the database, described by a feature combination of size P. 

Slip) is the value for feature p in the specification st. Likewise with Uj[p], 

y)is a function that returns the distance between x and y for the values of feature p. 

Hence the total target cost T is a Manhattan distance, calculated by summing P subcosts, one for each feature p in the unit and item feature structures. Another mechanism is to allow each sub-cost to be calculated using any subset of the feature structure, so the numbers of sub-costs and features are not necessarily equivalent. This, however, can be generalised to the above case by pre-processing the features. 

---