Simple Prediction of an Ecosystem-Specific Water Quality Index and the Water Quality Classification of a Highly Polluted River through Supervised Machine Learning

Water quality indices (WQIs) are used for the simple assessment and classification of water surface sources. However, considerable time, financial resources, effort required to measure parameters their calculation. Prediction WQIs through supervised machine learning is a useful approach reduce cost analysis development predictive models with reduced number parameters. In this study, regression machine-learning were developed estimate ecosystem-specific WQI previously Santiago-Guadalajara River (SGR-WQI), which involves measurement 17 The best subset selection method was employed significant SGR-WQI prediction. multiple linear model using 12 displayed residual square error (RSE) 3.262, similar that (RSE = 3.255), translates into savings estimation. Additionally, generalized additive not only an adjusted R2 0.9992, fit all evaluated, but also fitted rating curves each parameter original algorithm calculation great accuracy. Regarding models, overall proportion 93% 86% data correctly classified logistic parameters, respectively, while discriminant functions 84%. evaluated found be efficient in predicting as complementary tools extend current monitoring program River.