Search

This paper describes the feasibility of applying a combination of intelligent approaches of self organizing maps and fuzzy logic in modeling and predicting the growth of total bacteria coliform in tropical lakes. The application of self organizing maps (SOM) in this study is to derive the membership rules for ecological variables such as temperature, pH and biochemical oxygen demand (BOD) that influence the growth of bacteria coliform in the water of Putrajaya (tropical) Lakes and Wetlands. The membership rules derived from the SOM are used as the bases to tune the intelligent fuzzy prediction module. The overall system has been trained and tested for its accuracy using a reliable database that provides samples of relevant parameters from the lakes and wetlands over a period of five years. The system has demonstrated prediction of coliform growth with up to 90.5% of accuracy. [ABSTRACT FROM AUTHOR]