Do experts make mistakes? A comparison of human and machine identification of dinoflagellates

The authors present evidence of the difficulties facing human taxonomists/ecologists in identifying marine dinoflagellates. This is especially important for work on harmful algal blooms in marine aquaculture. It is shown that it is difficult for people to categorise specimens from species with significant morphological variation, perhaps with morphologies overlapping with those of other species. Trained personnel can be expected to achieve 67 to 83% self-consistency and 43% consensus between people in an expert taxonomic labelling task. Experts who are routinely engaged in particular discriminations can return accuracies in the range of 84 to 95%. In general, neither human nor machine can be expected to give highly accurate or repeatable labelling of specimens. It is also shown that automation methods can perform as well as humans on these complex categorisations.

[1]  R R Sokal,et al.  Classification: Purposes, Principles, Progress, Prospects , 1974, Science.

[2]  H. Perry Jeffries,et al.  COMPUTER-PROCESSING OF ZOOPLANKTON SAMPLES , 1980 .

[3]  Manfred Rolke,et al.  Size structure analysis of zooplankton samples by means of an automated image analyzing system , 1984 .

[4]  A. D. Poularikas,et al.  Automated sizing, counting and identification of zooplankton by pattern recognition , 1984 .

[5]  Jonathan S. Evans,et al.  Bias in human reasoning , 1990 .

[6]  Phil F. Culverhouse,et al.  Biological pattern recognition by neural networks , 1991 .

[7]  Mitochondrial DNA Variation of Copepods: Markers of Species Identity and Population Differentiation in Calanus. , 1991, The Biological bulletin.

[8]  Phil F. Culverhouse,et al.  Classification of euceratium gran. in neural networks , 1991, [1991 Proceedings] IEEE Conference on Neural Networks for Ocean Engineering.

[9]  Lynne Boddy,et al.  Neural Network Analysis of Flow Cytometry Data , 1993 .

[10]  Rob Williams,et al.  Speciation Of The Tintinnid Genus Cymatocylis By Morphometric Analysis Of The Loricae , 1994 .

[11]  F. Colijn,et al.  Phytoplankton monitoring by flow cytometry , 1994 .

[12]  Phil F. Culverhouse,et al.  Automatic categorisation of five species of Cymatocylis (Protozoa, Tintinnida) by artificial neural network , 1994 .

[13]  Flow cytometry: a powerful tool in analysis of biomass distributions in phytoplankton , 1994 .

[14]  Description of different morphotypes of D. acuminata complex in the Galician Rias Bajas in 1991 , 1995 .

[15]  Beatriz Reguera,et al.  Phytoplankton recognition using parametric discriminants , 1996 .

[16]  B. Reguera,et al.  Phased cell division and other biological observations in field populations of Dinophysis spp during cell cycle studies , 1996 .

[17]  A. Bucklin,et al.  The population genetics of Calanus finmarchicus in the North Atlantic , 1996 .

[18]  W. K. Stewart,et al.  Rapid visualization of plankton abundance and taxonomic composition using the Video Plankton Recorder , 1996 .

[19]  Loic Le Dean,et al.  Application of a digital pattern recognition system to Dinophysis acuminata and D. sacculus complexes , 1996 .

[20]  P. Truquet,et al.  Morphological Study of Atypical Dinophysis acuta Ehrenberg from Chilean Coastal Waters by a Digital Pattern - Recognition System , 1998 .

[21]  R. Hill,et al.  Multiplexed species-specific PCR protocol to discriminate four N. Atlantic Calanus species, with an mtCOI gene tree for ten Calanus species , 2001 .

[22]  A. Solow,et al.  Estimating the taxonomic composition of a sample when individuals are classified with error , 2001 .

[23]  S. González-Gil,et al.  A procedure to estimate okadaic acid in whole dinoflagellate cells using immunological techniques , 1995, Journal of Applied Phycology.

[24]  Phil F. Culverhouse,et al.  Committees, collectives and individuals: Expert visual classification by neural network , 1997, Neural Computing & Applications.