Efficient Drinking Water Quality Analysis using Machine Learning Model with Hyper-Parameter Tuning

The conventional approaches to determining the quality of water include costly and lengthy statistical and laboratory testing; hence, the concept of real-time monitoring is no longer applicable in the modern day. The disastrous implications of poor water quality demand that there be a remedy that is both more expedient and more realistic. This research study examines a variety of supervised machine learning algorithms in order to identify the quality of the water. A variety of variables, such as the pH value, the hardness, the solids, the chloramines, the sulfates, the conductivity, the organic carbon, the trihalomethanes, the turbidity, and the potability of the water, are essential for determining the quality of the water. Random Forest (RF) and Decision Tree (DT) are used in order to ascertain the caliber of the water that is suitable for human consumption. The standard laboratory approach for testing the quality of water is one that takes a lot of time and may be somewhat expensive at times. In a very short length of time, the algorithms that are proposed in this study are able to provide an estimation of the quality of drinking water. DT has a height accuracy F1 score of 99%, whereas RF has a score of 87.86% and a precision of 82.36%. The difference in these two scores is due to the fact that DT’s precision is lower. The suggested method demonstrates its potential for usage in real-time water quality monitoring systems by achieving appropriate accuracy with a minimal set of parameters. To prove the usefulness of this program, this is essential.