On the number of dependent variables for incompletely specified multiple-valued functions

This paper considers the minimization of dependent variables in functions with many don't cares. It also derives the conditions for almost all randomly generated function to be redundant in at least one variable. Experimental results support the validity of the approach.