Classification of Water Quality and Impurity Detection Using Machine Learning

Water quality assessment is considered one of the most necessary measures taken to safeguard public health and maintain ecological balance. This work aims to design a machine learning-based framework for determining the classification of water quality and impurity level based on key physical, chemical, and biological parameters. The dataset was collected from diversified water bodies across multiple regions and included measurements such as pH, Biological Oxygen Demand (BOD), conductivity, temperature, nitrate and nitrite concentration, and coliform bacteria levels. After scrubbing of data for sanitizing it, we had applied the machine learning model like Gradient Boosting which calculates an impurity measure about every place. The important sources of water pollution for a specified region are determined by making use of feature significance analysis. Power BI was used to further visualize the quality of water and prime factors driving the water quality in different regions. It generated an interactive map with dynamic visualisation of impurity hotspots across several places. Users can hover over specific places to get massive information about the key pollutants responsible for the deterioration of water quality. The results of the study point out certain areas that are contaminated with certain types of pollutants, including industrial areas that have higher BOD or nitrates that exceed the permissible limits in agricultural areas.