University of the Aegean
Βελτιστοποίηση μαθηματικών αλγορίθμων μάθησης στη θαλάσσια οικολογία
Abstract
dc:descriptionThe exploration of processes leading to coastal eutrophication is a major challenge in ecological research, particularly in light of important new policies such as the European Water Framework Directive. In the present study primary production, water quality status and phytoplankton diversity are modeled based on exclusively abiotic parameters using different machine learning techniques. Specifically, model trees showed increased predictive power in primary production prediction offering an explanatory description of ecosystem status. The water quality status was sufficiently classified using a voting ensemble method and a novel index was proposed in order to facilitate the optimization procedure during voting training. Finally, phytoplankton biodiversity was predicted in terms of its three components (richness, evenness and dominance) using both field and noise-free simulated assemblages. Based on the optimization of biodiversity prediction, a software package was developed for phytoplankton diversity prediction for Eastern Mediterranean waters.The study resulted in the development of information technology tools offering useful insights into ecosystem processes affecting eutrophication in coastal ecosystems, constituting also useful components in integrated coastal zone management. Moreover, the proposed methodologies can be easily extended or adapted to any group of organisms either in marine or terrestrial ecosystems. Possible future applications include also the incorporation of community structure in ecological models and global change scenarios.
Degree
thesis:*- Grantor dc:publisher
- University of the Aegean
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Tamvaki, Androniki
- Ταμβάκη, Ανδρονίκη
Subjects
dc:subject × 16- Μηχανική μάθηση
- Νευρωνικά δίκτυα
- Δένδρα Απόφασης
- Βιοποικιλότητα
- Περιεκτικότητα χλωροφύλης
- Συνδυαστικοί αλγόριθμοι βελτιστοποίησης
- Machine learning
- Artificial neural networks
- Decision trees
- Biodiversity
- Chlorophyll A
- Ensemble methods
- Φυσικές Επιστήμες
- Γεωεπιστήμες και Επιστήμες Περιβάλλοντος
- Natural Sciences
- Earth and Related Environmental Sciences
Rights
dc:rights- Statement dc:rights
-
- BY
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/34586