Abstracts

Diversity of the Fauna

In the analysis of animal comunity data, in particular on relationship of comunity characteristics from abiotic or anthropogenic factors, it is possible to use indexes. We use the premises offered by D.Brey: the best splitting of the factor on gradation will be if each species belongs to one gradation. If each species meets in all gradation, such splitting is assigned zero (Bioindication: theory, methods, applications, 1994).

The purpose of our activity is the calculation of indicator informativeness of animal comunity with fuzzy based approach. Our goal at the given phase have included formalisation of indicator informativeness of animal comunity, creating of algorithm of construction of this characteristic and constructing of a membership function of comunity informativeness of animal.

The data represent the array from n_grad of columns (number of factor’s gradation =5) and n_vid of rows (number of species = 8). If the species is absent then in column the miss of a species is denoted.

When the species meets not in all gradation, we obtain the formula of intermediate values of indicator informativeness of species:
IND=1 / (n_grad-1) * (n_grad -S), where n_grad - number of gradation of the factor, S - number of occurrings of species in various gradation of the factors.

At the fuzzy approach the linguistic variables are used instead of numerical variables and in addition to them, for example: {"small", "mean", "large"} (Zade, 1976; Averkin, Batyrshin, 1986).

The constructed estimations of indicator informativeness of animal comunity can serve for the characteristic of quality of splitting of the factors or bioindicator. It is possible to use estimations of plausibility obtained on the basis of a membership function, for construction of the rules in expert systems, for example: a rule R: If Terrain = " Panovsky forest " and indicator informativeness = large Than graduation of the factors =good, pv=4.Where pv=4 is plausibility values of rule. Plausibility values of rules take values from the set {0,1,2,3,4,5,6}. These numbers are correspondent to the verbal grades of the scale of plausibility values L:{maximal (6), very large (5), large (4), average (3), small (2), very small (1), minimal (0)}.

Note. Abstracts are published in author's edition

|Home Page| |English Part|

Go to Home