Robust function approximation under uncertainty

The paper investigates the properties of robustness (stability, insensitivity to noises) of optimal approximation of a function by algebraical polynomials with a modular criterion. Robustness conditions are found for the case of sampled-data noises.