A Comprehensive Review of Machine Learning Algorithms and Its Application in Groundwater Quality Prediction
Harsh Pandya,Khushi Jaiswal,Manan Shah
DOI: https://doi.org/10.1007/s11831-024-10126-2
IF: 9.7
2024-06-25
Archives of Computational Methods in Engineering
Abstract:Groundwater is among the utmost essential renewable resources for every organism existing on Earth. Assessing water quality is critical for the ecosystem's stability and conservation. The overall water quality possesses a significant effect on human being wellness and environmental preservation. Numerous applications of water exist, including those related to industries, agriculture, and consumption. The water quality index (WQI) is an essential metric for assessing water management effectiveness. By its biological, physical, and physiological features, water quality assesses whether water is suitable for a specific application or not. Water quality analysis has become a big concern in today's world because of industrialization, industry, farming techniques, and people's behavior. Quality of water has traditionally been examined using expensive testing facilities and numerical procedures, enabling monitoring in real-time obsolete. Improper quality of groundwater necessitates an additional feasible and affordable remedy. The algorithmic learning-based categorization technique looks to be promising for quick identification and estimation of water quality. Predicting the quality of water has been done effectively using machine learning algorithms. The technological investigation of computer algorithms as well as mathematical models that networks of computers employ to complete a certain task without having to be explicitly programmed is referred to as machine learning (ML). The major benefit associated with algorithmic machine learning models is that as an algorithm knows how to utilize data, it can perform its function independently. This work comprehensively examines three major machine learning techniques: Decision Tree, Regression Model, and Support Vector Machine. Features including total coliform, electric conductivity, biological oxygen demand, pH, dissolved oxygen, and nitrate determine the water quality. In this paper, many prior research that employed machine learning techniques for determining water quality in diverse regions were examined. A comparison of past research involving these algorithms, assessment methodologies, and acquired outcomes is offered. We performed a thorough analysis of the cutting-edge ML algorithms used to predict groundwater quality. As part of our methodology, we analysed a wide range of research, looked into the use of conventional and cutting-edge ML techniques, pre-processing techniques, feature selection techniques, and data augmentation methods. The findings of this study will help with groundwater development planning and will enhance the Machine learning applications in improving the quality of groundwater. Our analysis demonstrates the adaptability of ML techniques in predicting groundwater quality. We discovered that ML models, such as deep learning, ensemble approaches, neural networks, support vector machines, and linear regression, have been successfully used to predict the quality of groundwater, identify the origins of contamination, and optimise remediation techniques. We also point out how important data availability and quality are to model success.
computer science, interdisciplinary applications,engineering, multidisciplinary,mathematics