<![CDATA[Water is an essential role that need for human life, but not all water can be categorized as safe to drink. Therefore, it is compulsory to identify the classification between safe and unsafe water to drink. The purpose of this research is to determine the accuracy of water quality as much as 3,276 data with a pre-processing process to produce consistent data. In this study, the authors compare three methods in the data classification process, namely K-nearest neighbors, Naïve Bayes, and Decision Tree to find out the most accurate method with the maximum level of accuracy. The results showed that the highest level of accuracy is the K-nearest neighbors method with an accuracy rate of 71.19% where the class precision is for pred. zero (pred. negative) was 72.89%, pred. one (pred.positive) is 67.16%, while the Naïve Bayes method is 62.99% where the class precision is for pred. zero (pred. negative) is 64.26%, pred one (pred.positive) is 56.28%, and Decision tree is 61.77% where the class precision is for pred. zero (pred. negative) was 61.47%, pred one (pred.positive) was 100.00%.]]>
Copyrights © 2022
