Data from: Supervised classification of plant communities with artificial neural networks

This dataset was used to test the performance of artificial neural networks for supervised classification of plant communities, published in:
Černá L. & Chytrý M. (2005) Supervised classification of plant communities with artificial neural networks. Journal of Vegetation Science 16, 407-414. https://doi.org/10.1111/j.1654-1103.2005.tb02380.x
The meaning of the individual columns (separated by semicolons) in the file is as follows (for details see the above-mentioned article):

Plot no - unique number of the vegetation plot
Group expert - plot membership in classes 1-11 of the expert  classification
Subset expert random B - assignment of the plot to the training, selection, test or ignored data subset, using the random selection of the training (and selection) subset, for the expert classification
Subset expert dg species B - assignment of the plot to the training, selection, test or ignored data subset, using the selection of the training (and selection) subset by  diagnostic species, for the expert classification
Assignment expert random - a class assignment of the plot by the MLP classifier, when trained with the randomly selected training (and selection) subset, for expert classification
Assignment expert dg-sp - a class assignment  of the plot by the MLP classifier, when trained with the plots rich in diagnostic species contained in the training (and selection) subset, for expert classification
Group cluster - plot membership in classes 1-11 of the numerical classification
Subset cluster random - assignment of the plot to the training, selection, test or ignored data subset, using the random selection of the training (and selection) subset, for numerical classification
Subset cluster dg species - assignment of the plot to the training, selection, test or ignored data subset, using the selection of the training (and selection) subset by diagnostic species, for expert classification, for numerical classification
Assignment cluster random - class assignment of the plot by the MLP classifier, when trained with the randomly selected training (and selection) subset, for expert classification, for numerical classification
Assignment cluster dg-sp - class assignment  of the plot by the MLP classifier, when trained with the plots rich in diagnostic species contained in the training (and selection) subset, for expert classification, for numerical classification 
598 species, with cover/abundance estimates on an ordinal scale of 1-9