OUTLIERS IN STATISTICAL PATTERN-RECOGNITION AND AN APPLICATION TO AUTOMATIC CHROMOSOME CLASSIFICATION

We propose a heuristic method of parameter estimation in mixture model s for data with outliers and design a Bayesian classifier for assignme nt of m objects to n greater than or equal to m classes under constrai nts. This method of outlier handling combined with the classifier is a pplied to the well-known problem of automatic, constrained classificat ion of chromosomes into their biological classes. We show that it decr eases the error rate relative to the classical, normal, model by more than 50%. When applied to the Edinburgh feature data of the large Cope nhagen image data set Cpr our best classifier yields an error rate clo se to 1.3% relative to chromosomes; 4 out of 5 cells are correctly cla ssified. (C) 1997 Elsevier Science B.V.