The study’s objective was to develop a predictive model that classifies the turbidity of water using data mining techniques and aimed to help visualize data and predict the classification of the lake's water turbidity, whether it is good or bad. The parameters utilized in the study were Conductivity, Dissolved Oxygen (DO), pH, Total Suspended Solid (TSS), Total Coliform, and Temperature. Artificial Neural Network (ANN), Support Vector Machine (SVM), and k-Nearest Neighbor (KNN) are the data mining techniques used to create the models. The model's effectiveness tests for accuracy, precision, and recall. Correlation-based feature selection describes the linear relationship between different parameters and a model. The highest correlation was obtained between TSS and Turbidity among the attributes, while the temperature was the lowest. The study used three different combinations of parameters. The researchers found that the class count in the data affects the accuracy provided by the model. The less the count of one part of the binary classifier present in the data, the more likely the accuracy will be closer to one. The training of data was through the capabilities of Python. Laravel web framework used to develop the web-based application in PHP language. Furthermore, the results of high-quality development data are a foundation for meaningful insights to protect health and avoid water pollution in developing countries.