Use of self-organizing maps in modelling the distribution patterns of gammarids (Crustacea: Amphipoda)

Abstract

Self Organizing Maps (SOMs) are increasingly popular methods in processing highdimensional ecological data, however, their potentials are not yet fully utilized. It was our objective to prove evidence on an unknown advantage of the SOMs which we aimed to test using data on the spatial distributional patterns of gammarids. Quantitative samples and a wide spectrum of environmental data were obtained from the catchment area of two of the largest side tributaries of the Tisza River. Distributional patterns and habitat preference of three Gammarus species were described by Self Organizing Map methods and regression tree analysis (CART) on spatial and temporal scale. Using SOMs helped us to bring out distinctions in our data and enhance the differences, thus making them easier to recognize and A C C E P T E D M A N U S C R I P T ACCEPTED MANUSCRIPTalso, with their help, we were able to model the relations of the species to habitat types nonexistent among our samples. According to the analysis Gammarus roeselii preferred low altitudes, high conductivity, fine substrate, deep actual mean depth and dense plant coverage;G. fossarum preferred rocky stream beds, high altitude, lower temperatures and little actual mean depth; while G. balcanicus preferred coarse substrate, little or no plant coverage and low temperature. SOMs improved the correlations that proved to be highly useful: besides their use to display complex data in a perspicuous way they have other advantages in bringing out existing relationships in data otherwise difficult to detect.

Self-organizing maps for pattern recognition in design of alloys

“…Recent studies demonstrate the advantage of using SOM over PCA [23,26]. Most importantly, SOMs have been successfully used for feature extraction of scarce datasets (sample size of about 40), whereas conventional neural networks require large training sets [27,28]. In this work, we used a commercial optimization package mode FRONTIER for SOM analysis [5] which uses the following steps.…”

Reverse logistics risk management: identification, clustering and risk mitigation strategies

“…Länsiluoto and Eklund, 2008) and in other disciplines (e.g. Krasznai et al , 2016) which has adopted a similar approach with low sample size and yet achieved relatively good accuracy (see Table 5).…”

Design of High Temperature Ti–Al–Cr–V Alloys for Maximum Thermodynamic Stability Using Self-Organizing Maps

“…The SOM algorithm is a classification technique that is based upon an unsupervised artificial neural network, popularly known as self-organizing feature maps (SOM) [17][18][19][20][21][22], which was popularized by Teuvo Kohonen in the 1980s. SOM implements a term competitive learning along with a neighborhood function to preserve the topological properties of the dataset [21].…”

“…This makes SOM a perfect tool to visualize high-dimensional datasets in lower dimensions-usually two to three-while preserving the topology for determining various correlations within the dataset [18]. SOMs can be considered as a nonlinear generalization of principal component analysis (PCA), an unsupervised machine learning method [20,22]. Recent studies demonstrated the advantage of using SOM over PCA [10,14].…”