Dick, Daniel
Fuzzy Ecospace Modelling
<p>Fuzzy Ecospace Modelling (FEM) is an R-based program for quantifying and comparing functional disparity, using a fuzzy set theory-based machine learning approach. FEM clusters n-dimensional matrices of functional traits (ecospace matrices – here called the Training Matrix) into functional groups and converts them into fuzzy functional groups using fuzzy discriminant analysis (Lin and Chen 2004 – see main text for more information). Following this, FEM classifies the functional entities from a second matrix (the Test Matrix) into the groups made using the Training Matrix, generating fuzzy membership values for each unit in the Test Matrix. These values are real numbers from 0 to 1, representing increasing degrees of “truth” regarding an organism’s membership in the fuzzy set (see main text). A value of 0 represents non-membership in the fuzzy set, and a value of 1 represents total membership in the fuzzy set. Values in between represent degrees of niche overlap.</p>
<p>UPDATE: 2019-06-26: Fixed a common error which occured in Step 4E of the FEM algorithm.</p>
Ecology;Paleontology;Ecological Modelling
2019-07-01
https://zenodo.figshare.com/articles/Fuzzy_Ecospace_Modelling/8397557

10.5281/zenodo.3257750