Non-uniqueness in probabilistic numerical identification of bacteria
Authors:
- Mats GyllenbergMGMats Gyllenberg
- Timo KoskiTKTimo Koski
- Edwin ReilinkEREdwin Reilink
- Martin VerlaanMVMartin Verlaan
- 1 June 1994
- journal article
- Published by Cambridge University Press (CUP) in Journal of Applied Probability
- Vol. 31 (2) , 542-548
Abstract
In this note we point out an inherent difficulty in numerical identification of bacteria. The problem is that of uniqueness of the taxonomic structure or, in mathematical terms, the lack of statistical identifiability of finite mixtures of multivariate Bernoulli probability distributions shown here.Keywords
This publication has 28 references indexed in Scilit:
- Computer-assisted pattern recognition model for the identification of slowly growing mycobacteria including Mycobacterium tuberculosisJournal of General Microbiology, 1992
- Numerical classification and identification of Aeromonas genospeciesJournal of Applied Bacteriology, 1992
- Probabilistic identification of streptomycetes using miniaturized physiological testsJournal of General Microbiology, 1991
- A probability matrix for the identification of vibriosJournal of Applied Bacteriology, 1985
- A Probability Matrix for Identification of some StreptomycetesMicrobiology, 1983
- Identification of Bacteria by Computer: Identification of Reference StrainsJournal of General Microbiology, 1973
- Identification of Bacteria by Computer: General Aspects and PerspectivesJournal of General Microbiology, 1973
- Conditional Probability and the Identification of Bacteria: a Pilot StudyJournal of General Microbiology, 1968
- Paired Comparisons for Paired CharacteristicsThe Annals of Mathematical Statistics, 1968
- Estimating the Parameters of Mixtures of Binomial DistributionsJournal of the American Statistical Association, 1964