Water Quality Classification using XGBoost method

Water quality is threatened by a multitude of pollutants, which makes it impossible to provide humans with clean water. As a result, the importance of predicting and forecasting water quality is growing in order to reduce water pollution. Currently, expensive and time-consuming laboratory and statistical testing are used to evaluate the quality of water. In addition, machine learning techniques are used to evaluate the quality of water, such as SVM, KNN, and ANN. However, these techniques take longer to finish and are less accurate. In order to classify water quality, the tree-based XGBoost algorithm is utilized. One of a data set’s seven attributes, the Water Quality Index (WQI), is used to classify the WQC. A confusion matrix is used to assess this strategy’s effectiveness. With an accuracy rate of 94%, the analysis results showed that the XGBoost model performed better than the others. With XGBoost, just 6% of the data were misclassified. This approach is better since it can perform out-of-core calculations more quickly. Moreover, with an average accuracy of 90%, XGBoost continuously exceeded the fivefold validation. Because of its improved performance, it has been shown that XGBoost is more successful in improving the quality of the water.