WATER QUALITY PREDICTION USING LSTM-RNN: A REVIEW

• Published in: Journal of Sustainability Science and Management
• Main Author: Jaffar A.; Thamrin N.M.; Ali M.S.A.M.; Misnan M.F.; Yassin A.I.M.
• Format: Article
• Language: English
• Published: Universiti Malaysia Terengganu 2022
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85135834564&doi=10.46754%2fjssm.2022.07.015&partnerID=40&md5=7f32111ad0ed144efb4adb0e0dbf46ee

Water is a critical component of life on this planet, as humans, animals, and plants all rely on water supplies for survival. Regrettably, human activity has contaminated water sources. As a means of remedying this situation, each river should have an intelligent system that continuously monitors the water quality index at each water source. However, testing and establishing a realistic acceptability level for all water quality and control parameters takes time. Some of the data on the parameter’s values were also not immediately available. To address this problem, an estimation based on past samples and readings was applied. The process also implemented an exemplary computer system-based method known as an artificial neural network (ANN) Which has evolved into a promising and versatile tool for data prediction over the last few years. This article analyses and compares a host of ANN techniques that have been researched or used in the past to determine water quality. According to this study’s review of available literature, there are two types of ANN computation engines: Feedforward network systems and recurrent network systems that can be used for the purposes of determining water quality. Each method has advantages and disadvantages which were studied and this study hypothesises that with right techniques will be able to accurately predict and account for river water quality. © Penerbit UMT