Improve the Purity and Quality of the Water Using the Correlations Among the Feature Variables

EasyChair Preprint 9868, version 1

Abstract

The assessment of water purity and the position of sources for safe potable water has been one of the major problems the world has encountered recently. The primary challenge is preserving water and maintaining the purity of the water from extreme environmental pollution because the mode of water poisoning is unpredictable, making it difficult to assess and maintain. As a result, maximizing resource management is very efficient for improving water purity. There are some particular water-related factors required for the determination of the water quality indicator (WQI). However, the traditional way of computing the WQI requires a lot of effort, and occasionally mistakes are discovered during the computation procedure. In this research, we evaluate the cleanliness of the water and forecast its grade using machine learning techniques. If the water is out of purity our algorithms will identify the chemical compositions of the water and proposes with a required combination in order to make the water balance with pH and stable for drinking.Therefore, effective administration of water supplies is crucial to enhancing water purity, both conventionally and through the use of machine learning algorithms. This study's primary objectives are the collection of water data, analysis of chemical combinations, correlations between combinations, water samples were analyzed for their physico-chemical characteristics, the WQI was calculated using this information, and machine learning methods were used to forecast water purity. Several statistical metrics were used to evaluate the success of the forecast algorithms.

Keyphrases: WQI, Water purity, chemical compositions, correlations, machine learning

@booklet{EasyChair:9868,
  author    = {L Roshini and C.R.K. Reddy},
  title     = {Improve the Purity and Quality of the Water Using the Correlations Among the Feature Variables},
  howpublished = {EasyChair Preprint 9868},
  year      = {EasyChair, 2023}}