In L p -spaces with p.elem.(0,1] there exists a best approximation mapping to the set of functions computable by Heaviside perceptron networks with n hidden units; however for p.elem.(0,1) such best approximation is not unique and cannot be continuous.
In L p -spaces with p.elem.(0,1] there exists a best approximation mapping to the set of functions computable by Heaviside perceptron networks with n hidden units; however for p.elem.(0,1) such best approximation is not unique and cannot be continuous. (en)