查看文章

This study examines the use of an intelligent system to predict the quality of water flowing through the Johore Rivers. A neural network model of multilayer perceptron has been developed to predict the quality of water in three rivers, namely the Bekok River, Sayong River, and Johor River. The model has been trained and tested with the values of the pH parameters as input variables. The neural network model’s performance has been assessed using three performance metrics: mean squared error, CPU time, and prediction accuracy. By analyzing the relationship between the residuals and the model values, it will be possible to assess the model’s suitability. Based on the results from the implementation of the multilayer perceptron model, a method is developed for forecasting and monitoring the parameters associated with water quality obtaining an average accuracy of prediction score of 99.93%.