(Received August 3, 1990; revised June 13, 1991)
Abstract. We study the choice of the quantity alpha in the FPEalpha criterion for selecting a member of a class of normal linear models having an orthogonal structure. Two approaches are discussed, namely fixing the maximal estimation risk at a prescribed level and using minimax regret. Estimation of the risk actually achieved and an illustrative example are also discussed.
Key words and phrases: Model selection, linear model, estimation, FPEalpha criterion, minimax regret, risk estimation.