Enhancing Accuracy in Water Pollution Prediction that Results in Chronic Infections on Marine Life Using Multi-linear Regression and Random Forest Regression

Predicting water pollution accuracy is a critical aspect of everyday life because water quality severely impacted by chemicals released from industrial waste, sewage, bacteria, plastics, etc. Pollutants directly harm fish populations, causing chronic diseases and death. This study focuses on predicting water pollution accurately using machine learning algorithms like multi-linear regression and random forest regression. Samples were collected from various water bodies across different states in India. The dataset was processed through preprocessing, normalization, water quality index calculation, and modeling using multi-linear regression and random forest regression. Performance was evaluated using mean square and root mean square. Predicted results were compared with the SPSS tool using independent variable test, and outcomes were graphically represented with a bar chart. The findings indicate that random forest regression outperforms multi-linear regression in accurately predicting water quality and fish disease. This research assists policymakers at state level in making effective decisions to control water pollution.