Trophic Diatom Classification Using Naгџve Bayes

Naumovski, A. and Mitreski, Kosta (2013) Trophic Diatom Classification Using Naгџve Bayes. In: Proceedings of the Tenth Conference on Informatics and Information Technology. Faculty of Computer Science and Engineering, Ss. Cyril and Methodius University in Skopje, Macedonia, Skopje, Macedonia, pp. 131-135. ISBN 978-608-4699-01-9

[img]
Preview
Text
978-608-4699-01-9_pp131-135.pdf

Download (322kB) | Preview
Official URL: http://ciit.finki.ukim.mk

Abstract

Knowledge discovery has been used in many different type of analysis and data types which lead to increased understanding of many natural processes and phenomena. This is why this process is important in the area of analysing environmental data. The topic and the goal of the paper is to used this process and the information contained in the measured data for given lake ecosystem and extract that information in an understandable form. This research aims to assess the relationships between the diatoms and the indicators of the environment using Naïve Bayes method learning technique. The diatoms are taken into account because they are ideal indicators of certain physical-chemical parameters and they can be classified into one of the trophic quality classes (TQCs). Before the algorithm processes the data, the input dataset is discretised. Then using the Naïve Bayes technique, several models for each TQC are obtained, presented and discussed. Then the obtained knowledge is verified with existing diatom ecological preference. Directions of future research and improvement for using this method for environmental data are given at the conclusion of the paper.

Item Type: Book Section
Subjects: International Conference on Informatics and Information Technologies > Intelligent Systems
International Conference on Informatics and Information Technologies > Robotics
International Conference on Informatics and Information Technologies > Bioinformatics
Depositing User: Vangel Ajanovski
Date Deposited: 28 Oct 2016 00:15
Last Modified: 28 Oct 2016 00:15
URI: http://eprints.finki.ukim.mk/id/eprint/11011

Actions (login required)

View Item View Item