STOCHASTIC-MODELS APPLIED TO CLUSTER-ANALYSIS OF CENSORED WATER-QUALITY DATA

A clustering method is presented which is appropriate when a mixture o f multivariate distributions is used to model the data. The procedure assumes that the measurements made on several independent characterist ics follow probability distributions of different nature (e.g., normal , Poisson, multinomial). Likelihood methods are used to both estimate the parameters of the models and perform the cluster analysis. The pro cedure is applicable when the data are subject to censoring. A data se t on the concentrations of volatile halocarbon contaminant in Lake Ont ario is used to illustrate the methods of this paper.