Enhanced Potable Water Detection Using Feature Fusion and Selection with Extreme Learning Machine

Accurately determining the potability of water is crucial because access to clean, safe drinking water is crucial for maintaining public health. Thus, to enhance the prediction results outlined in this work, we introduced employing deep learning in the rule-based classification of the water potability system using an efficient machine learning algorithm. The first step taken with the data involved handling data discrepancy while the second involved scaling of the data known as data normalization. The tools applied to decide which of the water quality variables has the greatest impact on the model results include, ANOVA, Information Gain, and Gain Ratio. Following that, with the help of these improved characteristics, the Extreme Learning Machine (ELM) gets a classification rate of 90%. The suggested model intends to enhance the monitoring and decision-making regarding water quality enhancement since the method offered will help in potable water assessment.