Least squares regression of the Fuzzy linear model is extended to over
come and interpret the occurrence of negative spreads tl]. The idea is
to introduce best fit difference models which exploit both the Hukuha
ra difference and the L-2-metric distance. The fuzzy models use LR-fuz
zy numbers. Fitted models are compared by using the coefficient of det
ermination, in a similar way to its use in classical statistical least
squares fitting. The non-LR-fuzzy case is also considered.