This paper describes the feasibility study of applying a hybrid combination of Kohonen self organizing feature maps (SOM) and a rule based system in predicting the biomass of selected algae division (Chlorophyta) at tropical Putrajaya Lake (Malaysia). The system was trained and tested on an over five years of limnological time-series data sampled from Putrajaya Lake. Results from trained SOM were used to extract rules of relationships between input variables and the Chlorophyta biomass which was used to construct a rule based system. Selected input variables were water temperature, Secchi depth and nitrate nitrogen (NO3-N). The rules extracted conformed to findings as postulated in literatures. The overall rule based system yielded an accuracy of 73%.
Key words: Kohonen self organizing feature maps, prediction, rule based system, chlorophyta.
Copyright © 2024 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0