Please use this identifier to cite or link to this item: http://hdl.handle.net/20.500.12188/22686
Title: Multi-target modelling of the diatom diversity indices in Lake Prespa
Authors: Naumoski, Andreja 
Keywords: diatoms, Multi-target modelling, Lake Prespa, machine learning
Issue Date: 1-Jan-2012
Publisher: CORVINUS UNIV BUDAPEST
Journal: Applied Ecology and Environmental Research
Abstract: In this paper we present models of relationship between the diatoms community diversity indices (DIs) and the physico-chemical parameters using machine learning techniques. By taking relative abundances into account, a diversity index depends not only on species richness but also on the evenness, or equitability, with which individuals are distributed among the different time and space. Diversity indices provide important information about rarity and commonness of species in a community. Because the physical-chemical conditions of the environmental influence on the several diversity indices of the diatoms community at once, it is more reliably to model all the diversity indices together. For modelling of the DIs models we use the raw; as measured, values of the concentrations for the physical-chemical parameters and the diversity indices of the diatoms abundance. The well known machine learning techniques are used to express this relationship: regression trees (RTs) and multi-target regression trees (MTRT’s). The MTRT are more general than the RT, which predictive target is only one variable. The diversity indices are calculated for all diatoms of one measurement for 16 months, monthly and then are placed with the given physico-chemical parameters in one table. The results from the model have captured the ecological information with correlation between 0.9 and 0.92 for unseen (test) data. Diversity indices have proved to be a reliable indicator for the influence of the environment on the diatoms community. Temperature and conductivity components together with the Zn oncentration are most influenced factors on the diatoms biodiversity. This could lead to more widely research broad view in this direction of ecological modelling.
URI: http://hdl.handle.net/20.500.12188/22686
Appears in Collections:Faculty of Computer Science and Engineering: Journal Articles

Files in This Item:
File Description SizeFormat 
EPA02583_applied_ecology_2012_04_521-529.pdf136.08 kBAdobe PDFView/Open
Show full item record

Page view(s)

38
checked on Sep 22, 2024

Download(s)

10
checked on Sep 22, 2024

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.