BAYESIAN RANK ESTIMATION WITH APPLICATION TO FACTOR-ANALYSIS

Rank estimation is a common sub-problem met in various fields exploiti ng matrix algebra. Estimation of the number of factors in factor analy sis is of this type. Due to large noise contents in the analyzed matri x, standard procedures, like deterministic inspection of singular valu es, fail. In the paper, a novel procedure is proposed. It is gained by straightforward application of Bayesian statistics to a carefully sel ected model which fits to the target area, namely, factor analysis of dynamic scintigraphic studies. The formal solution consists of an exac tly feasible part and a maximum-likelihood type one. The latter is jus tified by large dimensions of the data matrices containing analyzed im ages. Properties of the procedure are illustrated on simulated and rea l data.