This paper deals with an "inverse problem": the estimation of an initial distribution in the first boundary value problem for the heat equation through some biased information on its solution. Numerically stable solutions to the inverse problem are normally achieved through various regularization procedures. It is shown that these procedures could be treated within a unified framework of solving guaranteed estimation problems for systems with unknown but bounded errors.