Decision tree models in predicting water quality parameters of dissolved oxygen and phosphorus in lake water

Water quality is an important issue because of its relationship to humans and other living organisms. Predicting water quality parameters is very important for better management of water resources. The decision tree is one of the data mining methods that can create rules for classifying and predicting data using a tree structure. The purpose of this study is to use data mining techniques to investigate and predict the parameters of soluble phosphorus and oxygen in Lake Erie to achieve this purpose. The Classification And Regression Tree (CART) model is compared with the Chi-squared Automatic Interaction Detector (CHAID) model and the Quick Unbiased Efficient Statistical Trees (QUEST) model with the C5 model. Comparison and review of these models to express their applicability to identify water quality parameters are conducted. The results show that decision tree methods with the help of hydrochemical parameters can classify and predict water quality with high accuracy and in a short time. The number of available data is 327. To check the accuracy of the models, the difference between the observed data and the predicted data is used. In the prediction of dissolved oxygen, 214 cases with the CART model and 185 cases with the CHAID model differ by less than 2 units from the observed data. For phosphorus, 245 cases in the CART model and 237 cases in the CHAID model differ less than 0.2 the predicted data with the observed data. Therefore, the accuracy of the CART model is better. The prediction of 256 phosphorus parameter group numbers and 230 dissolved oxygen parameter group numbers with the C5 algorithm is correct. The results show that CART model is better than CHAID model in predicting data, and C5 model is better than QUEST model in predicting group numbers.