Machine Learning Approach for Water Purity Detection

Abstract

Humans rely on resources from the earth, which are valued for their quality. Climate change and its effects on the environment have long been monitored. The product\'s quality is carefully assessed before to usage. Water, an unavoidable resource, has become increasingly important in terms of quality monitoring due to the influence of different external elements such as industrial effluents, acid rain, and so on. This research presents a system for monitoring water quality that employs statistical quality control techniques and machine learning algorithms to improve the accuracy of classification. Here we are using gradient boosting, Random forest, XG boost, naïve bayes, Ada boost, cat boost algorithms the categorization is concerned with determining whether the water is appropriate for home use or not.

Introduction

The text discusses the importance of predicting drinking water quality using machine learning techniques to ensure safe and healthy water consumption. Contaminated water can cause serious health problems, making accurate and timely water quality assessment essential. Traditional testing methods are often expensive, time-consuming, and unable to provide real-time predictions. Machine learning provides an efficient alternative by analyzing water parameters and predicting whether water is safe for consumption.

The quality of drinking water is determined using parameters such as pH, hardness, sulfate, organic carbon, turbidity, solids, conductivity, chloramines, and trihalomethanes. Various machine learning algorithms, including Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN), Deep Neural Networks (DNN), Random Forest (RF), and Support Vector Machines (SVM), have been applied for water quality prediction.

The literature review highlights previous research using machine learning models such as Random Forest, Support Vector Machine, Artificial Neural Networks, Decision Trees, K-Nearest Neighbors, Logistic Regression, Gradient Boosting, and deep learning methods for assessing water purity. These studies emphasize the importance of feature selection, preprocessing, and early warning systems for detecting water contamination.

The existing water quality assessment methods face challenges due to complex mathematical modeling, lengthy computations, and difficulties in handling large datasets. To overcome these limitations, the proposed system applies machine learning algorithms from the scikit-learn library for efficient classification of potable and non-potable water.

The proposed system uses multiple classification algorithms, including:

• Random Forest
• Gradient Boosting
• Gaussian Naive Bayes
• XGBoost
• AdaBoost
• CatBoost

The methodology involves collecting the Kaggle Water Potability Dataset, which contains 3277 records and 10 features. The dataset includes water characteristics such as pH, hardness, dissolved solids, chloramines, sulfate, conductivity, organic carbon, trihalomethanes, turbidity, and a potability label. Data preprocessing steps include:

• Handling missing values
• Removing duplicate data
• Feature scaling and normalization
• Splitting data into training (70%) and testing (30%) sets

The trained machine learning models are evaluated based on their prediction accuracy to identify the most effective algorithm for water quality classification. Algorithms such as Random Forest use multiple decision trees for improved prediction, AdaBoost combines weak learners into a stronger classifier, XGBoost provides efficient gradient boosting for large datasets, and Naive Bayes performs probability-based classification.

The results demonstrate that machine learning models can effectively classify water as potable or non-potable, providing faster and more reliable predictions compared with conventional testing methods. The study concludes that machine learning-based water quality prediction systems can support environmental monitoring, pollution prevention, and public health protection by enabling accurate and efficient drinking water assessment.

Conclusion

In summary, studying weather forecasts using Predicting usable water is critical for environmental sustainability and pollution avoidance. Clean drinking water is essential for maintaining good public health. Drinking water from secure sources ensures that the water is portable. It becomes harder to make reliable predictions about drinking water. The perfect learning algorithm is required to avoid prediction errors. An intelligent model based on five different machine learning algorithms can be used to forecast the portability of drinking water using ten standard criteria like as pH, hardness, organic carbon, and other variables.

References

[1] S. Geetha, S. Gouthami. Internet of Things Enabled Real Time Water Quality Monitoring System ,2017. [2] Umair Ahmed, Rafia Mumtaz, Hirra Anwar, Asad A. Shah, Rabia Irfan and José García-Nieto. Efficient Water Quality Prediction Using Supervised Machine Learning, 2019. [3] Ashwini K, D. Diviya, J.Janice Vedha, M. Deva Priya.Intelligent Model For Predicting Water Quality. [4] A.N.Prasad, K. A. Mamun, F. R. Islam, H. Haqva.Smart Water Quality Monitoring System, 2015. [5] Hadi Mohammed, Ibrahim A. Hameed, Razak Seidu.Machine Learning: Based Detection of Water Contamination in Water Distribution systems,2018. [6] Priya Singh,Pankaj Deep Kaur.Review on Data Mining Techniques for Prediction of Water Quality,2017. [7] Manish Kumar Jha,Rajni Kumari Sah,M.S.Rashmitha,Rupam Sinha,B.Sujatha.Smart Water Monitoring System for Real-Time Water Quality and Usage Monitoring,2018 . [8] Water QualityMonitoring System using IoT and Machine Learning, in Proceedings of the IEEE International Conference onResearch in Intelligent and Computing in Engineering, pp.1-5, 2018. [9] Amir Hamzeh Haghiabi, Ali Heidar Nasrolahi, Abbas Parsaie. Water quality prediction using machine learning methods, 2018.

Copyright

Copyright © 2026 Y. Hari , Tulasi Miriyala. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.