ARIMA Modeling for Reliable Potable Water Identification and Quality Prediction

This work aims to employ machine learning methods and visualization techniques for evaluating water quality by measuring its potability, a numerical indicator of water quality. The work utilizes water quality parameters to determine the stability condition using $\mathbf{p H}$, Temperature, Dissolved Oxygen, Conductivity, Nitrate, Fecal Coliform, and Total Coliform to evaluate the overall water quality. ARIMA is utilized to identify and categorize water potability based on its features. Using Tableau’s interactive dashboarding, we created an interactive dashboard to visually simplify water potability metrics in India. The study utilized both real datasets from various locations around the country using government data sourced directly from the Central Pollution Control Board (CPCB). Based on the results of the experiments, the ARIMA was found to be superior to other classifiers in accurately predicting water potability.