Generalizations

Distortion (a type of generalization): suppose instead of f(x) we report a random number from a distribution Df
Example: choose a value y uniformly in the interval [f(x)-a, f(x)+a]
Suppose we consider f(x) identified if |y - f(x)|  
Properties of f and D may enable us to bound f(z) (a different value) definitively
Properties of f and D may enable us to bound f(z) probabilistically

Distortion (a type of generalization): suppose instead of f(x) we report a random number from a distribution Df