Classification of Drinking Water Potability With Artificial Neural Network Algorithm

Abstract

Having safe water for consumption is essential for public health in every region. However, water quality is declining in some places, especially to meet human needs for drinking water. There are many efforts to maintain water potability, such as checking to see if there are bacteria or diseases in the water. This research classifies water potability using the Artificial Neural Network method, a technique in the field of machine learning. This research classifies water quality using a python library to analyze data and perform classification. Data is processed through stages such as data cleaning and data division into training and testing. In testing, the data is divided into 20% for testing and 80% for training. The results of the ANN algorithm show 70% accuracy. in conclusion, the ANN model has moderate performance in classifying the feasibility of drinking water. Model improvement is needed to improve accuracy and prediction, including the use of larger and more diverse datasets.